[pre-commit.ci] pre-commit autoupdate (#10183)

updates:
- [github.com/astral-sh/ruff-pre-commit: v0.1.7 → v0.1.8](https://github.com/astral-sh/ruff-pre-commit/compare/v0.1.7...v0.1.8)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

.github/workflows/benchmarks.yaml

@@ -10,7 +10,7 @@ jobs:
     name: run benchmarks
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/setup-python@v4
+      - uses: actions/setup-python@v5
         with:
             python-version: "3.11"
       - uses: iterative/dvc-bench@main

.github/workflows/build.yaml

@@ -23,7 +23,7 @@ jobs:
       with:
         fetch-depth: 0
 
-    - uses: actions/setup-python@v4
+    - uses: actions/setup-python@v5
       with:
         python-version: 3.8
 
@@ -39,7 +39,7 @@ jobs:
         SETUPTOOLS_SCM_PRETEND_VERSION_FOR_DVC: ${{ env.version }}
     - run: twine check --strict dist/*
 
-    - uses: actions/upload-artifact@v3
+    - uses: actions/upload-artifact@v4
       with:
         name: Packages
         path: dist/
@@ -58,7 +58,7 @@ jobs:
       id-token: write
 
     steps:
-    - uses: actions/download-artifact@v3
+    - uses: actions/download-artifact@v4
       with:
         name: Packages
         path: dist
@@ -83,7 +83,7 @@ jobs:
       id-token: write
 
     steps:
-    - uses: actions/download-artifact@v3
+    - uses: actions/download-artifact@v4
       with:
         name: Packages
         path: dist

.github/workflows/codeql.yml

@@ -31,9 +31,10 @@ jobs:
       uses: actions/checkout@v4
 
     - name: Initialize CodeQL
-      uses: github/codeql-action/init@v2
+      uses: github/codeql-action/init@v3
       with:
         languages: ${{ matrix.language }}
+        queries: security-extended
 
     - name: Perform CodeQL Analysis
-      uses: github/codeql-action/analyze@v2
+      uses: github/codeql-action/analyze@v3

.github/workflows/plugin_tests.yaml

@@ -34,7 +34,7 @@ jobs:
         path: ${{ matrix.plugin }}
 
     - name: Set up Python
-      uses: actions/setup-python@v4
+      uses: actions/setup-python@v5
       with:
         python-version: ${{ matrix.pyv }}
         cache: pip

.github/workflows/tests.yaml

@@ -23,14 +23,14 @@ jobs:
       fail-fast: false
       matrix:
         os: [ubuntu-20.04, windows-latest, macos-latest]
-        pyv: ["3.8", "3.9", "3.10", "3.11"]
+        pyv: ["3.8", "3.9", "3.10", "3.11", "3.12"]
 
     steps:
     - uses: actions/checkout@v4
       with:
         fetch-depth: 0
     - name: Set up Python
-      uses: actions/setup-python@v4
+      uses: actions/setup-python@v5
       with:
         python-version: ${{ matrix.pyv }}
         cache: pip

.pre-commit-config.yaml

@@ -22,7 +22,7 @@ repos:
       - id: sort-simple-yaml
       - id: trailing-whitespace
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: 'v0.1.5'
+    rev: 'v0.1.8'
     hooks:
       - id: ruff
         args: [--fix, --exit-non-zero-on-fix]

dvc/_debug.py

@@ -23,10 +23,11 @@ def viztracer_profile(
     tracer = viztracer.VizTracer(max_stack_depth=depth, log_async=log_async)
 
     tracer.start()
-    yield
-    tracer.stop()
-
-    tracer.save(path() if callable(path) else path)
+    try:
+        yield
+    finally:
+        tracer.stop()
+        tracer.save(path() if callable(path) else path)
 
 
 @contextmanager
@@ -45,31 +46,33 @@ def yappi_profile(
     yappi.set_clock_type("wall" if wall_clock else "cpu")
 
     yappi.start()
-    yield
-    yappi.stop()
+    try:
+        yield
+    finally:
+        yappi.stop()
 
-    threads = yappi.get_thread_stats()
-    stats = {}
-    if separate_threads:
-        for thread in threads:
-            ctx_id = thread.id
-            stats[ctx_id] = yappi.get_func_stats(ctx_id=ctx_id)
-    else:
-        stats[None] = yappi.get_func_stats()
-
-    fpath = path() if callable(path) else path
-    for ctx_id, st in stats.items():
-        if fpath:
-            out = f"{fpath}-{ctx_id}" if ctx_id is not None else fpath
-            st.save(out, type="callgrind")
+        threads = yappi.get_thread_stats()
+        stats = {}
+        if separate_threads:
+            for thread in threads:
+                ctx_id = thread.id
+                stats[ctx_id] = yappi.get_func_stats(ctx_id=ctx_id)
         else:
-            if ctx_id is not None:
-                print(f"\nThread {ctx_id}")  # noqa: T201
-            st.print_all()
-            if ctx_id is None:
-                threads.print_all()
+            stats[None] = yappi.get_func_stats()
 
-    yappi.clear_stats()
+        fpath = path() if callable(path) else path
+        for ctx_id, st in stats.items():
+            if fpath:
+                out = f"{fpath}-{ctx_id}" if ctx_id is not None else fpath
+                st.save(out, type="callgrind")
+            else:
+                if ctx_id is not None:
+                    print(f"\nThread {ctx_id}")  # noqa: T201
+                st.print_all()
+                if ctx_id is None:
+                    threads.print_all()
+
+        yappi.clear_stats()
 
 
 @contextmanager
@@ -85,13 +88,15 @@ def instrument(html_output=False):
     profiler = Profiler()
 
     profiler.start()
-    yield
-    profiler.stop()
+    try:
+        yield
+    finally:
+        profiler.stop()
 
-    if html_output:
-        profiler.open_in_browser()
-        return
-    print(profiler.output_text(unicode=True, color=True))  # noqa: T201
+        if html_output:
+            profiler.open_in_browser()
+        else:
+            print(profiler.output_text(unicode=True, color=True))  # noqa: T201
 
 
 @contextmanager
@@ -102,13 +107,14 @@ def profile(dump_path: Optional[str] = None):
     prof = cProfile.Profile()
     prof.enable()
 
-    yield
-
-    prof.disable()
-    if not dump_path:
-        prof.print_stats(sort="cumtime")
-        return
-    prof.dump_stats(dump_path)
+    try:
+        yield
+    finally:
+        prof.disable()
+        if dump_path:
+            prof.dump_stats(dump_path)
+        else:
+            prof.print_stats(sort="cumtime")
 
 
 @contextmanager

dvc/cachemgr.py

@@ -27,7 +27,7 @@ def _get_odb(
     cls, config, fs_path = get_cloud_fs(repo.config, **settings)
     fs = fs or cls(**config)
     if prefix:
-        fs_path = fs.path.join(fs_path, *prefix)
+        fs_path = fs.join(fs_path, *prefix)
     if hash_name:
         config["hash_name"] = hash_name
     return get_odb(fs, fs_path, state=repo.state, **config)
@@ -86,7 +86,7 @@ class CacheManager:
         return FileStorage(
             key=(),
             fs=self.local.fs,
-            path=self.local.fs.path.join(self.default_local_cache_dir, self.FS_DIR),
+            path=self.local.fs.join(self.default_local_cache_dir, self.FS_DIR),
         )
 
     def _init_odb(self, schemes):

dvc/cli/parser.py

@@ -19,6 +19,7 @@ from dvc.commands import (
     data_sync,
     destroy,
     diff,
+    du,
     experiments,
     freeze,
     gc,
@@ -26,6 +27,7 @@ from dvc.commands import (
     get_url,
     git_hook,
     imp,
+    imp_db,
     imp_url,
     init,
     install,
@@ -41,6 +43,7 @@ from dvc.commands import (
     repro,
     root,
     stage,
+    studio,
     unprotect,
     update,
     version,
@@ -65,6 +68,7 @@ COMMANDS = [
     data_sync,
     gc,
     imp,
+    imp_db,
     imp_url,
     config,
     checkout,
@@ -91,6 +95,8 @@ COMMANDS = [
     check_ignore,
     data,
     artifacts,
+    studio,
+    du,
 ]
 
 

dvc/commands/config.py

@@ -1,6 +1,8 @@
 import argparse
 import os
 
+from funcy import set_in
+
 from dvc.cli.command import CmdBaseNoRepo
 from dvc.cli.utils import append_doc_link
 from dvc.log import logger
@@ -8,7 +10,7 @@ from dvc.ui import ui
 
 logger = logger.getChild(__name__)
 
-NAME_REGEX = r"^(?P<remote>remote\.)?(?P<section>[^\.]*)\.(?P<option>[^\.]*)$"
+NAME_REGEX = r"^(?P<top>(remote|db)\.)?(?P<section>[^\.]*)\.(?P<option>[^\.]*)$"
 
 
 def _name_type(value):
@@ -17,10 +19,12 @@ def _name_type(value):
     match = re.match(NAME_REGEX, value)
     if not match:
         raise argparse.ArgumentTypeError(
-            "name argument should look like remote.name.option or section.option"
+            "name argument should look like remote.name.option or "
+            "db.name.option or section.option"
         )
+    top = match.group("top")
     return (
-        bool(match.group("remote")),
+        top.strip(".") if top else None,
         match.group("section").lower(),
         match.group("option").lower(),
     )
@@ -49,12 +53,11 @@ class CmdConfig(CmdBaseNoRepo):
             logger.error("name argument is required")
             return 1
 
-        remote, section, opt = self.args.name
+        remote_or_db, section, opt = self.args.name
 
         if self.args.value is None and not self.args.unset:
-            return self._get(remote, section, opt)
-
-        return self._set(remote, section, opt)
+            return self._get(remote_or_db, section, opt)
+        return self._set(remote_or_db, section, opt)
 
     def _list(self):
         if any((self.args.name, self.args.value, self.args.unset)):
@@ -75,18 +78,18 @@ class CmdConfig(CmdBaseNoRepo):
 
         return 0
 
-    def _get(self, remote, section, opt):
+    def _get(self, remote_or_db, section, opt):
         from dvc.config import ConfigError
 
         levels = self._get_appropriate_levels(self.args.level)[::-1]
 
         for level in levels:
             conf = self.config.read(level)
-            if remote:
-                conf = conf["remote"]
+            if remote_or_db:
+                conf = conf[remote_or_db]
 
             try:
-                self._check(conf, remote, section, opt)
+                self._check(conf, remote_or_db, section, opt)
             except ConfigError:
                 if self.args.level:
                     raise
@@ -99,16 +102,15 @@ class CmdConfig(CmdBaseNoRepo):
 
         return 0
 
-    def _set(self, remote, section, opt):
+    def _set(self, remote_or_db, section, opt):
         with self.config.edit(self.args.level) as conf:
-            if remote:
-                conf = conf["remote"]
+            if remote_or_db:
+                conf = conf[remote_or_db]
             if self.args.unset:
-                self._check(conf, remote, section, opt)
+                self._check(conf, remote_or_db, section, opt)
                 del conf[section][opt]
             else:
-                self._check(conf, remote, section)
-                conf[section][opt] = self.args.value
+                conf.update(set_in(conf, [section, opt], self.args.value))
 
         if self.args.name == "cache.type":
             logger.warning(
@@ -119,10 +121,10 @@ class CmdConfig(CmdBaseNoRepo):
 
         return 0
 
-    def _check(self, conf, remote, section, opt=None):
+    def _check(self, conf, remote_or_db, section, opt=None):
         from dvc.config import ConfigError
 
-        name = "remote" if remote else "section"
+        name = remote_or_db or "section"
         if section not in conf:
             raise ConfigError(f"{name} '{section}' doesn't exist")
 

dvc/commands/dag.py

@@ -86,7 +86,7 @@ def _collect_targets(repo, target, outs):
             targets.extend([str(out) for out in stage.outs])
             continue
 
-        for out in outs_trie.itervalues(prefix=repo.fs.path.parts(path)):
+        for out in outs_trie.itervalues(prefix=repo.fs.parts(path)):
             targets.extend(str(out))
 
     return targets

dvc/commands/du.py

@@ -0,0 +1,80 @@
+import argparse
+import logging
+
+from dvc.cli import completion
+from dvc.cli.command import CmdBaseNoRepo
+from dvc.cli.utils import DictAction, append_doc_link
+from dvc.ui import ui
+
+logger = logging.getLogger(__name__)
+
+
+class CmdDU(CmdBaseNoRepo):
+    def run(self):
+        from dvc.repo import Repo
+        from dvc.utils.humanize import naturalsize
+
+        entries = Repo.du(
+            self.args.url,
+            self.args.path,
+            rev=self.args.rev,
+            summarize=self.args.summarize,
+            config=self.args.config,
+            remote=self.args.remote,
+            remote_config=self.args.remote_config,
+        )
+        ui.table([(naturalsize(size), path) for path, size in entries])
+        return 0
+
+
+def add_parser(subparsers, parent_parser):
+    DU_HELP = "Show disk usage."
+    du_parser = subparsers.add_parser(
+        "du",
+        parents=[parent_parser],
+        description=append_doc_link(DU_HELP, "du"),
+        help=DU_HELP,
+        formatter_class=argparse.RawTextHelpFormatter,
+    )
+    du_parser.add_argument("url", help="Location of DVC repository")
+    du_parser.add_argument(
+        "--rev",
+        nargs="?",
+        help="Git revision (e.g. SHA, branch, tag)",
+        metavar="<commit>",
+    )
+    du_parser.add_argument(
+        "-s",
+        "--summarize",
+        action="store_true",
+        help="Show total disk usage.",
+    )
+    du_parser.add_argument(
+        "--config",
+        type=str,
+        help=(
+            "Path to a config file that will be merged with the config "
+            "in the target repository."
+        ),
+    )
+    du_parser.add_argument(
+        "--remote",
+        type=str,
+        help="Remote name to set as a default in the target repository.",
+    )
+    du_parser.add_argument(
+        "--remote-config",
+        type=str,
+        nargs="*",
+        action=DictAction,
+        help=(
+            "Remote config options to merge with a remote's config (default or one "
+            "specified by '--remote') in the target repository."
+        ),
+    )
+    du_parser.add_argument(
+        "path",
+        nargs="?",
+        help="Path to directory within the repository",
+    ).complete = completion.DIR
+    du_parser.set_defaults(func=CmdDU)

dvc/commands/imp_db.py

@@ -0,0 +1,167 @@
+import argparse
+
+from funcy import compact, merge
+
+from dvc.cli import completion
+from dvc.cli.command import CmdBase
+from dvc.cli.utils import append_doc_link
+from dvc.log import logger
+from dvc.ui import ui
+
+logger = logger.getChild(__name__)
+
+
+class CmdTestDb(CmdBase):
+    def run(self):
+        from dvc.database import get_client
+        from dvc.database.dbt_utils import DBT_PROJECT_FILE, is_dbt_project
+        from dvc.dependency.db import _get_dbt_config
+        from dvc.exceptions import DvcException
+
+        connection = self.args.connection
+
+        db_config = self.repo.config.get("db", {})
+        config = db_config.get(connection, {})
+        if connection and not config:
+            raise DvcException(f"connection {connection} not found in config")
+
+        cli_config = compact(
+            {
+                "url": self.args.url,
+                "username": self.args.username,
+                "password": self.args.password,
+            }
+        )
+        conn_config = merge(config, cli_config)
+
+        cli_dbt_config = {"profile": self.args.profile, "target": self.args.target}
+        dbt_config = merge(_get_dbt_config(self.repo.config), cli_dbt_config)
+
+        project_dir = self.repo.root_dir
+        if not (conn_config or dbt_config):
+            if is_dbt_project(project_dir):
+                ui.write("Using", DBT_PROJECT_FILE, "for testing", styled=True)
+            else:
+                raise DvcException(
+                    f"no config set and {DBT_PROJECT_FILE} is missing; "
+                    "provide arguments or set a configuration"
+                )
+
+        adapter = get_client(conn_config, project_dir=project_dir, **dbt_config)
+        with adapter as db:
+            ui.write(f"Testing with {db}", styled=True)
+
+            creds = getattr(db, "creds", {})
+            for k, v in creds.items():
+                ui.write("\t", f"{k}:", v, styled=True)
+
+            if creds:
+                ui.write()
+
+            db.test_connection()
+
+        ui.write("Connection successful", styled=True)
+
+
+class CmdImportDb(CmdBase):
+    def run(self):
+        if not (self.args.sql or self.args.model):
+            raise argparse.ArgumentTypeError("Either of --sql or --model is required.")
+
+        self.repo.imp_db(
+            url=self.args.url,
+            rev=self.args.rev,
+            project_dir=self.args.project_dir,
+            sql=self.args.sql,
+            model=self.args.model,
+            profile=self.args.profile,
+            target=self.args.target,
+            output_format=self.args.output_format,
+            out=self.args.out,
+            force=self.args.force,
+            connection=self.args.connection,
+        )
+        return 0
+
+
+def add_parser(subparsers, parent_parser):
+    IMPORT_HELP = (
+        "Download file or directory tracked by DVC or by Git "
+        "into the workspace, and track it."
+    )
+
+    import_parser = subparsers.add_parser(
+        "import-db",
+        parents=[parent_parser],
+        description=append_doc_link(IMPORT_HELP, "import"),
+        add_help=False,
+    )
+    import_parser.add_argument(
+        "--url", help="Location of DVC or Git repository to download from"
+    )
+    import_parser.add_argument(
+        "--rev",
+        nargs="?",
+        help="Git revision (e.g. SHA, branch, tag)",
+        metavar="<commit>",
+    )
+    import_parser.add_argument(
+        "--project-dir", nargs="?", help="Subdirectory to the dbt project location"
+    )
+
+    group = import_parser.add_mutually_exclusive_group()
+    group.add_argument(
+        "--sql",
+        help="SQL query",
+    )
+    group.add_argument(
+        "--model",
+        help="Model name to download",
+    )
+    import_parser.add_argument("--profile", help="Profile to use")
+    import_parser.add_argument("--target", help="Target to use")
+    import_parser.add_argument(
+        "--output-format",
+        default="csv",
+        const="csv",
+        nargs="?",
+        choices=["csv", "json"],
+        help="Export format",
+    )
+    import_parser.add_argument(
+        "-o",
+        "--out",
+        nargs="?",
+        help="Destination path to download files to",
+        metavar="<path>",
+    ).complete = completion.FILE
+    import_parser.add_argument(
+        "-f",
+        "--force",
+        action="store_true",
+        default=False,
+        help="Override destination file or folder if exists.",
+    )
+    import_parser.add_argument(
+        "--conn",
+        dest="connection",
+        nargs="?",
+        help="Database connection to use, needs to be set in config",
+    )
+
+    import_parser.set_defaults(func=CmdImportDb)
+
+    TEST_DB_HELP = "Test the database connection"
+    test_db_parser = subparsers.add_parser(
+        "test-db",
+        parents=[parent_parser],
+        description=append_doc_link(TEST_DB_HELP, "test-db"),
+        add_help=False,
+    )
+    test_db_parser.add_argument("--conn", dest="connection")
+    test_db_parser.add_argument("--url")
+    test_db_parser.add_argument("--password")
+    test_db_parser.add_argument("--username")
+    test_db_parser.add_argument("--profile")
+    test_db_parser.add_argument("--target")
+    test_db_parser.set_defaults(func=CmdTestDb)

dvc/commands/plots.py

@@ -59,49 +59,6 @@ def _show_json(
     ui.write_json(compact({"errors": all_errors, "data": data}), highlight=False)
 
 
-def _adjust_vega_renderers(renderers):
-    from dvc.render import REVISION_FIELD, VERSION_FIELD
-    from dvc_render import VegaRenderer
-
-    for r in renderers:
-        if isinstance(r, VegaRenderer):
-            if _data_versions_count(r) > 1:
-                summary = _summarize_version_infos(r)
-                for dp in r.datapoints:
-                    vi = dp.pop(VERSION_FIELD, {})
-                    keys = list(vi.keys())
-                    for key in keys:
-                        if not (len(summary.get(key, set())) > 1):
-                            vi.pop(key)
-                    if vi:
-                        dp["rev"] = "::".join(vi.values())
-            else:
-                for dp in r.datapoints:
-                    dp[REVISION_FIELD] = dp[VERSION_FIELD]["revision"]
-                    dp.pop(VERSION_FIELD, {})
-
-
-def _summarize_version_infos(renderer):
-    from collections import defaultdict
-
-    from dvc.render import VERSION_FIELD
-
-    result = defaultdict(set)
-
-    for dp in renderer.datapoints:
-        for key, value in dp.get(VERSION_FIELD, {}).items():
-            result[key].add(value)
-    return dict(result)
-
-
-def _data_versions_count(renderer):
-    from itertools import product
-
-    summary = _summarize_version_infos(renderer)
-    x = product(summary.get("filename", {None}), summary.get("field", {None}))
-    return len(set(x))
-
-
 class CmdPlots(CmdBase):
     def _func(self, *args, **kwargs):
         raise NotImplementedError
@@ -175,11 +132,10 @@ class CmdPlots(CmdBase):
                 return 0
 
             renderers = [r.renderer for r in renderers_with_errors]
-            _adjust_vega_renderers(renderers)
             if self.args.show_vega:
                 renderer = first(filter(lambda r: r.TYPE == "vega", renderers))
                 if renderer:
-                    ui.write_json(renderer.get_filled_template(as_string=False))
+                    ui.write_json(renderer.get_filled_template())
                 return 0
 
             output_file: Path = (Path.cwd() / out).resolve() / "index.html"

dvc/commands/studio.py

@@ -0,0 +1,175 @@
+import argparse
+import os
+
+from funcy import get_in
+
+from dvc.cli.utils import append_doc_link, fix_subparsers
+from dvc.commands.config import CmdConfig
+from dvc.log import logger
+
+logger = logger.getChild(__name__)
+
+
+class CmdStudioLogin(CmdConfig):
+    def run(self):
+        from dvc_studio_client.auth import StudioAuthError, get_access_token
+
+        from dvc.env import DVC_STUDIO_URL
+        from dvc.ui import ui
+        from dvc.utils.studio import STUDIO_URL
+
+        name = self.args.name
+        hostname = self.args.hostname or os.environ.get(DVC_STUDIO_URL) or STUDIO_URL
+        scopes = self.args.scopes
+
+        try:
+            token_name, access_token = get_access_token(
+                token_name=name,
+                hostname=hostname,
+                scopes=scopes,
+                use_device_code=self.args.no_open,
+                client_name="DVC",
+            )
+        except StudioAuthError as e:
+            ui.error_write(str(e))
+            return 1
+
+        self.save_config(hostname, access_token)
+        ui.write(
+            "Authentication has been successfully completed."
+            "The generated token will now be accessible as"
+            f" {token_name} in the user's Studio profile."
+        )
+        return 0
+
+    def save_config(self, hostname, token):
+        with self.config.edit("global") as conf:
+            conf["studio"]["token"] = token
+            conf["studio"]["url"] = hostname
+
+
+class CmdStudioLogout(CmdConfig):
+    def run(self):
+        from dvc.ui import ui
+
+        with self.config.edit("global") as conf:
+            if not get_in(conf, ["studio", "token"]):
+                ui.error_write("Not logged in to Studio.")
+                return 1
+
+            del conf["studio"]["token"]
+
+        ui.write("Logged out from Studio")
+        return 0
+
+
+class CmdStudioToken(CmdConfig):
+    def run(self):
+        from dvc.ui import ui
+
+        conf = self.config.read("global")
+        token = get_in(conf, ["studio", "token"])
+        if not token:
+            ui.error_write("Not logged in to Studio.")
+            return 1
+
+        ui.write(token)
+        return 0
+
+
+def add_parser(subparsers, parent_parser):
+    STUDIO_HELP = "Commands to authenticate DVC with Iterative Studio"
+    STUDIO_DESCRIPTION = (
+        "Authenticate DVC with Studio and set the token."
+        " Once this token has been properly configured,\n"
+        " DVC will utilize it for seamlessly sharing live experiments\n"
+        " and sending notifications to Studio regarding any experiments"
+        " that have been pushed."
+    )
+
+    studio_parser = subparsers.add_parser(
+        "studio",
+        parents=[parent_parser],
+        description=append_doc_link(STUDIO_DESCRIPTION, "studio"),
+        help=STUDIO_HELP,
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+    studio_subparser = studio_parser.add_subparsers(
+        dest="cmd",
+        help="Use `DVC studio CMD --help` to display command-specific help.",
+    )
+    fix_subparsers(studio_subparser)
+
+    STUDIO_LOGIN_HELP = "Authenticate DVC with Studio host"
+    STUDIO_LOGIN_DESCRIPTION = (
+        "By default, this command authenticates the DVC with Studio\n"
+        " using default scopes and assigns a random name as the token name."
+    )
+    login_parser = studio_subparser.add_parser(
+        "login",
+        parents=[parent_parser],
+        description=append_doc_link(STUDIO_LOGIN_DESCRIPTION, "studio/login"),
+        help=STUDIO_LOGIN_HELP,
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+
+    login_parser.add_argument(
+        "-H",
+        "--hostname",
+        action="store",
+        default=None,
+        help="The hostname of the Studio instance to authenticate with.",
+    )
+    login_parser.add_argument(
+        "-s",
+        "--scopes",
+        action="store",
+        default=None,
+        help="The scopes for the authentication token. ",
+    )
+
+    login_parser.add_argument(
+        "-n",
+        "--name",
+        action="store",
+        default=None,
+        help="The name of the authentication token. It will be used to\n"
+        "identify token shown in Studio profile.",
+    )
+
+    login_parser.add_argument(
+        "--no-open",
+        action="store_true",
+        default=False,
+        help="Use authentication flow based on user code.\n"
+        "You will be presented with user code to enter in browser.\n"
+        "DVC will also use this if it cannot launch browser on your behalf.",
+    )
+    login_parser.set_defaults(func=CmdStudioLogin)
+
+    STUDIO_LOGOUT_HELP = "Logout user from Studio"
+    STUDIO_LOGOUT_DESCRIPTION = (
+        "This removes the studio token from your global config.\n"
+    )
+
+    logout_parser = studio_subparser.add_parser(
+        "logout",
+        parents=[parent_parser],
+        description=append_doc_link(STUDIO_LOGOUT_DESCRIPTION, "studio/logout"),
+        help=STUDIO_LOGOUT_HELP,
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+
+    logout_parser.set_defaults(func=CmdStudioLogout)
+
+    STUDIO_TOKEN_HELP = "View the token dvc uses to contact Studio"  # noqa: S105 # nosec B105
+
+    logout_parser = studio_subparser.add_parser(
+        "token",
+        parents=[parent_parser],
+        description=append_doc_link(STUDIO_TOKEN_HELP, "studio/token"),
+        help=STUDIO_TOKEN_HELP,
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+
+    logout_parser.set_defaults(func=CmdStudioToken)

dvc/config.py

@@ -14,7 +14,7 @@ from .utils.objects import cached_property
 
 if TYPE_CHECKING:
     from dvc.fs import FileSystem
-    from dvc.types import DictStrAny, StrPath
+    from dvc.types import DictStrAny
 
 logger = logger.getChild(__name__)
 
@@ -89,8 +89,8 @@ class Config(dict):
 
     def __init__(
         self,
-        dvc_dir: Optional["StrPath"] = None,
-        local_dvc_dir: Optional["StrPath"] = None,
+        dvc_dir: Optional[str] = None,
+        local_dvc_dir: Optional[str] = None,
         validate: bool = True,
         fs: Optional["FileSystem"] = None,
         config: Optional["DictStrAny"] = None,
@@ -105,7 +105,7 @@ class Config(dict):
         self.fs = fs or self.wfs
 
         if dvc_dir:
-            self.dvc_dir = self.fs.path.abspath(dvc_dir)
+            self.dvc_dir = self.fs.abspath(dvc_dir)
 
         self.local_dvc_dir = local_dvc_dir
         if not fs and not local_dvc_dir:
@@ -145,10 +145,10 @@ class Config(dict):
         }
 
         if self.dvc_dir is not None:
-            files["repo"] = self.fs.path.join(self.dvc_dir, self.CONFIG)
+            files["repo"] = self.fs.join(self.dvc_dir, self.CONFIG)
 
         if self.local_dvc_dir is not None:
-            files["local"] = self.wfs.path.join(self.local_dvc_dir, self.CONFIG_LOCAL)
+            files["local"] = self.wfs.join(self.local_dvc_dir, self.CONFIG_LOCAL)
 
         return files
 
@@ -302,11 +302,11 @@ class Config(dict):
             return path.def_path
 
         if os.path.expanduser(path) != path:
-            return localfs.path.as_posix(path)
+            return localfs.as_posix(path)
 
         if isinstance(path, RelPath) or not os.path.isabs(path):
             path = relpath(path, conf_dir)
-            return localfs.path.as_posix(path)
+            return localfs.as_posix(path)
 
         return path
 
@@ -390,10 +390,10 @@ class Config(dict):
 
 
 def _parse_named(conf):
-    result: Dict[str, Dict] = {"remote": {}, "machine": {}}
+    result: Dict[str, Dict] = {"remote": {}, "machine": {}, "db": {}}
 
     for section, val in conf.items():
-        match = re_find(r'^\s*(remote|machine)\s*"(.*)"\s*$', section)
+        match = re_find(r'^\s*(remote|machine|db)\s*"(.*)"\s*$', section)
         if match:
             key, name = match
             result[key][name] = val
@@ -408,7 +408,7 @@ def _pack_named(conf):
     result = compact(conf)
 
     # Transform remote.name -> 'remote "name"'
-    for key in ("remote", "machine"):
+    for key in ("remote", "machine", "db"):
         for name, val in conf[key].items():
             result[f'{key} "{name}"'] = val
         result.pop(key, None)

dvc/config_schema.py

@@ -318,6 +318,8 @@ SCHEMA = {
     "feature": FeatureSchema(
         {
             Optional("machine", default=False): Bool,
+            "dbt_profile": str,
+            "dbt_target": str,
         },
     ),
     "plots": {
@@ -350,4 +352,11 @@ SCHEMA = {
         "repo_url": str,
         Optional("offline", default=False): Bool,
     },
+    "db": {
+        str: {
+            "url": str,
+            "username": str,
+            "password": str,
+        },
+    },
 }

dvc/data_cloud.py

@@ -3,7 +3,6 @@
 from typing import TYPE_CHECKING, Iterable, Optional, Set, Tuple
 
 from dvc.config import NoRemoteError, RemoteConfigError
-from dvc.fs.callbacks import Callback
 from dvc.log import logger
 from dvc.utils.objects import cached_property
 from dvc_data.hashfile.db import get_index
@@ -36,11 +35,9 @@ class Remote:
 
         path = self.path
         if self.worktree:
-            path = self.fs.path.join(
-                path, ".dvc", CacheManager.FILES_DIR, DEFAULT_ALGORITHM
-            )
+            path = self.fs.join(path, ".dvc", CacheManager.FILES_DIR, DEFAULT_ALGORITHM)
         else:
-            path = self.fs.path.join(path, CacheManager.FILES_DIR, DEFAULT_ALGORITHM)
+            path = self.fs.join(path, CacheManager.FILES_DIR, DEFAULT_ALGORITHM)
         return get_odb(self.fs, path, hash_name=DEFAULT_ALGORITHM, **self.config)
 
     @cached_property
@@ -206,11 +203,13 @@ class DataCloud:
         jobs: Optional[int] = None,
         odb: "HashFileDB",
     ) -> "TransferResult":
+        from dvc.fs.callbacks import TqdmCallback
+
         if odb.hash_name == "md5-dos2unix":
             cache = self.repo.cache.legacy
         else:
             cache = self.repo.cache.local
-        with Callback.as_tqdm_callback(
+        with TqdmCallback(
             desc=f"Pushing to {odb.fs.unstrip_protocol(odb.path)}",
             unit="file",
         ) as cb:
@@ -265,11 +264,13 @@ class DataCloud:
         jobs: Optional[int] = None,
         odb: "HashFileDB",
     ) -> "TransferResult":
+        from dvc.fs.callbacks import TqdmCallback
+
         if odb.hash_name == "md5-dos2unix":
             cache = self.repo.cache.legacy
         else:
             cache = self.repo.cache.local
-        with Callback.as_tqdm_callback(
+        with TqdmCallback(
             desc=f"Fetching from {odb.fs.unstrip_protocol(odb.path)}",
             unit="file",
         ) as cb:

dvc/database/__init__.py

@@ -0,0 +1,22 @@
+from typing import Any, ContextManager, Optional, Union
+
+from . import dbt_query, sqla
+from .dbt_models import get_model
+from .serializer import export
+
+Client = Union[sqla.SQLAlchemyClient, dbt_query.DbtClient]
+
+
+def get_client(
+    config,
+    project_dir: Optional[str] = None,
+    profile: Optional[str] = None,
+    target: Optional[str] = None,
+    **kwargs: Any,
+) -> "ContextManager[Client]":
+    if config:
+        return sqla.client(config, **kwargs)
+    return dbt_query.client(project_dir=project_dir, profile=profile, target=target)
+
+
+__all__ = ["export", "get_client", "get_model", "Client"]

dvc/database/dbt_models.py

@@ -0,0 +1,40 @@
+from typing import List, Optional
+
+from dvc.log import logger
+
+from .dbt_runner import dbt_show
+from .dbt_utils import check_dbt
+from .serializer import AgateSerializer
+
+logger = logger.getChild(__name__)
+
+
+def ref(
+    name: str,
+    package: Optional[str] = None,
+    version: Optional[int] = None,
+) -> str:
+    parts: List[str] = []
+    if package:
+        parts.append(repr(package))
+
+    parts.append(repr(name))
+    if version:
+        parts.append(f"{version=}")
+
+    inner = ",".join(parts)
+    return "{{ ref(" + inner + ") }}"
+
+
+@check_dbt("model")
+def get_model(
+    name: str,
+    package: Optional[str] = None,
+    version: Optional[int] = None,
+    profile: Optional[str] = None,
+    target: Optional[str] = None,
+) -> AgateSerializer:
+    model = ref(name, package, version=version)
+    q = f"select * from {model}"  # noqa: S608
+    table = dbt_show(inline=q, profile=profile, target=target)
+    return AgateSerializer(table)

dvc/database/dbt_query.py

@@ -0,0 +1,88 @@
+import os
+from contextlib import contextmanager
+from typing import TYPE_CHECKING, Any, Callable, Dict, Iterator, Optional, Union
+
+from attrs import define, field
+
+from dvc.exceptions import DvcException
+from dvc.log import logger
+
+from .dbt_utils import check_dbt
+from .serializer import AgateSerializer, PandasSQLSerializer
+
+if TYPE_CHECKING:
+    from dbt.adapters.base.impl import BaseAdapter
+
+logger = logger.getChild(__name__)
+
+
+@define
+class DbtClient:
+    adapter: "BaseAdapter" = field(repr=lambda o: type(o).__qualname__)
+    creds: Dict[str, Any] = field(repr=False)
+
+    @contextmanager
+    def query(self, sql: str) -> Iterator[Union[AgateSerializer, PandasSQLSerializer]]:
+        from dbt.adapters.sql import SQLAdapter
+
+        with self.adapter.connection_named("execute"):
+            if isinstance(self.adapter, SQLAdapter):
+                # try to use connection to stream results
+                # otherwise fallback to fetch all at once
+                try:
+                    connection = self.adapter.connections.get_thread_connection().handle
+                except Exception:  # noqa: BLE001
+                    logger.debug("", exc_info=True)
+                else:
+                    # pandas raises warning that it does not support
+                    # dbapi2-based raw connection, but it does seem to work.
+                    yield PandasSQLSerializer(sql, connection)
+                    return
+            _, table = self.adapter.execute(sql, fetch=True)
+            yield AgateSerializer(table)
+
+    def test_connection(self, onerror: Optional[Callable[[], Any]] = None) -> None:
+        with self.adapter.connection_named("debug"):
+            try:
+                self.adapter.debug_query()
+            except Exception as exc:
+                if callable(onerror):
+                    onerror()
+
+                logger.exception(
+                    "dvc was unable to connect to the specified database. "
+                    "Please check your database credentials and try again.",
+                    exc_info=False,
+                )
+                raise DvcException("The database returned the following error") from exc
+
+
+@contextmanager
+@check_dbt("query")
+def client(
+    project_dir: Optional[str] = None,
+    profiles_dir: Optional[str] = None,
+    profile: Optional[str] = None,
+    target: Optional[str] = None,
+) -> Iterator["DbtClient"]:
+    from dbt.adapters import factory as adapters_factory
+    from dbt.adapters.base.impl import BaseAdapter
+
+    from .dbt_utils import get_or_build_profile, get_profiles_dir, init_dbt
+
+    profiles_dir = profiles_dir or get_profiles_dir(project_dir)
+    flags = init_dbt(profiles_dir, os.getcwd(), target=target)
+
+    with flags, adapters_factory.adapter_management():
+        # likely invalid connection profile or no adapter
+        profile_obj = get_or_build_profile(project_dir, profile, target)
+
+        adapters_factory.register_adapter(profile_obj)  # type: ignore[arg-type]
+        adapter = adapters_factory.get_adapter(profile_obj)  # type: ignore[arg-type]
+        assert isinstance(adapter, BaseAdapter)
+        try:
+            creds = dict(profile_obj.credentials.connection_info())
+        except:  # noqa: E722
+            creds = {}
+
+        yield DbtClient(adapter, creds)

dvc/database/dbt_runner.py

@@ -0,0 +1,61 @@
+from typing import TYPE_CHECKING, List, Optional, Sequence, Union
+
+from dvc.exceptions import DvcException
+from dvc.log import logger
+
+if TYPE_CHECKING:
+    from agate import Table
+    from dbt.contracts.results import RunResult
+
+logger = logger.getChild(__name__)
+
+
+class DbtInternalError(DvcException):
+    pass
+
+
+def _kw_to_cmd_args(**kwargs: Union[None, bool, str, int]) -> List[str]:
+    args: List[str] = []
+    for key, value in kwargs.items():
+        key = key.replace("_", "-")
+        if value is None:
+            continue  # skip creating a flag in this case
+        if value is True:
+            args.append(f"--{key}")
+        elif value is False:
+            args.append(f"--no-{key}")
+        else:
+            args.extend([f"--{key}", str(value)])
+    return args
+
+
+def _dbt_invoke(*posargs: str, quiet: bool = True, **kw: Union[None, bool, str, int]):
+    from dbt.cli.main import dbtRunner
+
+    args = _kw_to_cmd_args(quiet=quiet or None)  # global options
+    args.extend([*posargs, *_kw_to_cmd_args(**kw)])
+
+    runner = dbtRunner()
+    result = runner.invoke(args)
+    if result.success:
+        return result.result
+    raise DbtInternalError(f"failed to run dbt {posargs[0]}") from result.exception
+
+
+def dbt_show(
+    inline: Optional[str] = None,
+    limit: int = -1,
+    profile: Optional[str] = None,
+    target: Optional[str] = None,
+) -> "Table":
+    from dbt.contracts.results import RunExecutionResult
+
+    result = _dbt_invoke(
+        "show", inline=inline, limit=limit, profile=profile, target=target
+    )
+    assert isinstance(result, RunExecutionResult)
+
+    run_results: Sequence["RunResult"] = result.results
+    run_result, *_ = run_results
+    assert run_result.agate_table is not None
+    return run_result.agate_table

dvc/database/dbt_utils.py

@@ -0,0 +1,100 @@
+import os
+from contextlib import contextmanager
+from importlib.util import find_spec
+from typing import TYPE_CHECKING, Optional
+
+from funcy import cut_prefix
+
+from dvc.exceptions import DvcException
+from dvc.log import logger
+from dvc.utils import packaging
+
+if TYPE_CHECKING:
+    from dbt.config.profile import Profile
+
+logger = logger.getChild(__name__)
+
+DBT_PROJECT_FILE = "dbt_project.yml"
+
+
+@contextmanager
+def check_dbt(action: Optional[str] = None):
+    if not (find_spec("dbt") and find_spec("dbt.cli")):
+        action = f" {action}" if action else ""
+        raise DvcException(f"Could not run{action}. dbt-core is not installed")
+    yield packaging.check_required_version(pkg="dbt-core")
+
+
+@contextmanager
+def init_dbt(
+    profiles_dir: str,
+    project_dir: str,
+    target: Optional[str] = None,
+):
+    from argparse import Namespace
+
+    from dbt import flags
+
+    prev = flags.get_flags()
+    try:
+        args = Namespace(
+            use_colors=True,
+            project_dir=project_dir,
+            profiles_dir=profiles_dir,
+            target=target,
+        )
+        flags.set_from_args(args, None)
+        yield flags.get_flags()
+    finally:
+        flags.set_flags(prev)
+
+
+def get_profiles_dir(project_dir: Optional[str] = None) -> str:
+    from dbt.cli.resolvers import default_profiles_dir
+
+    if profiles_dir := os.getenv("DBT_PROFILES_DIR"):
+        return profiles_dir
+    if project_dir and os.path.isfile(os.path.join(project_dir, "profiles.yml")):
+        return project_dir
+    return os.fspath(default_profiles_dir())
+
+
+@contextmanager
+def _handle_profile_parsing_error():
+    from dbt.exceptions import DbtRuntimeError
+
+    try:
+        yield
+    except DbtRuntimeError as e:
+        cause = e.__cause__ is not None and e.__cause__.__context__
+        if isinstance(cause, ModuleNotFoundError) and (
+            adapter := cut_prefix(cause.name, "dbt.adapters.")
+        ):
+            # DbtRuntimeError is very noisy, so send it to debug
+            logger.debug("", exc_info=True)
+            raise DvcException(f"dbt-{adapter} dependency is missing") from cause
+        raise DvcException("failed to read connection profiles") from e
+
+
+def is_dbt_project(project_dir: str):
+    return os.path.isfile(os.path.join(project_dir, DBT_PROJECT_FILE))
+
+
+@_handle_profile_parsing_error()
+def get_or_build_profile(
+    project_dir: Optional[str], profile: Optional[str], target: Optional[str]
+) -> "Profile":
+    from dbt.config.profile import Profile
+    from dbt.config.renderer import ProfileRenderer
+    from dbt.config.runtime import load_profile
+
+    if project_dir and is_dbt_project(project_dir):
+        return load_profile(
+            project_dir, {}, profile_name_override=profile, target_override=target
+        )
+
+    if not profile:
+        raise DvcException("No profile specified to query from.")
+
+    renderer = ProfileRenderer({})
+    return Profile.render(renderer, profile, target_override=target)

dvc/database/serializer.py

@@ -0,0 +1,54 @@
+from dataclasses import dataclass
+from typing import TYPE_CHECKING, Union
+
+if TYPE_CHECKING:
+    import sqlalchemy as sa
+    from agate import Table
+
+
+def noop(_):
+    pass
+
+
+@dataclass
+class PandasSQLSerializer:
+    sql: "Union[sa.TextClause, str]"
+    con: "sa.Connection"
+    chunksize: int = 10_000
+
+    def to_csv(self, file: str, progress=noop) -> None:
+        import pandas as pd
+
+        with open(file, mode="wb") as f:
+            idfs = pd.read_sql_query(self.sql, self.con, chunksize=self.chunksize)
+            for i, df in enumerate(idfs):
+                df.to_csv(f, header=i == 0, index=False)
+                progress(len(df))
+
+    def to_json(self, file: str, progress=noop) -> None:  # noqa: ARG002
+        import pandas as pd
+
+        df = pd.read_sql_query(self.sql, self.con)
+        df.to_json(file, orient="records")
+
+
+@dataclass
+class AgateSerializer:
+    table: "Table"
+
+    def to_csv(self, file: str, progress=noop) -> None:  # noqa: ARG002
+        return self.table.to_csv(file)
+
+    def to_json(self, file: str, progress=noop) -> None:  # noqa: ARG002
+        return self.table.to_json(file)
+
+
+def export(
+    serializer: Union[PandasSQLSerializer, AgateSerializer],
+    file: str,
+    format: str = "csv",  # noqa: A002
+    progress=noop,
+) -> None:
+    if format == "csv":
+        return serializer.to_csv(file, progress=progress)
+    return serializer.to_json(file, progress=progress)

dvc/database/sqla.py

@@ -0,0 +1,73 @@
+from contextlib import contextmanager
+from dataclasses import dataclass
+from typing import Any, Callable, Dict, Iterator, Optional, Union
+
+from sqlalchemy import create_engine, text
+from sqlalchemy.engine import URL, Engine
+from sqlalchemy.engine import make_url as _make_url
+from sqlalchemy.exc import NoSuchModuleError
+
+from dvc.exceptions import DvcException
+from dvc.log import logger
+
+from .serializer import PandasSQLSerializer
+
+logger = logger.getChild(__name__)
+
+
+def make_url(url: Union[URL, str], **kwargs: Any) -> URL:
+    return _make_url(url).set(**kwargs)
+
+
+def url_from_config(config: Union[str, URL, Dict[str, str]]) -> URL:
+    if isinstance(config, (str, URL)):
+        return make_url(config)
+    return make_url(**config)
+
+
+@dataclass
+class SQLAlchemyClient:
+    engine: Engine
+
+    @contextmanager
+    def query(self, sql: str) -> Iterator[PandasSQLSerializer]:
+        with self.engine.connect().execution_options(stream_results=True) as conn:
+            yield PandasSQLSerializer(text(sql), conn)
+
+    def test_connection(self, onerror: Optional[Callable[[], Any]] = None) -> None:
+        try:
+            with self.engine.connect() as conn:
+                conn.execute(text("select 1"))
+        except Exception as exc:
+            if callable(onerror):
+                onerror()
+            logger.exception(
+                "dvc was unable to connect to the specified database. "
+                "Please check your database credentials and try again.",
+                exc_info=False,
+            )
+            raise DvcException("The database returned the following error") from exc
+
+
+@contextmanager
+def handle_error(url: URL):
+    try:
+        yield
+    except (ModuleNotFoundError, NoSuchModuleError) as e:
+        # TODO: write installation instructions
+        raise DvcException(
+            f"Could not load database driver for {url.drivername!r}"
+        ) from e
+
+
+@contextmanager
+def client(
+    url_or_config: Union[str, URL, Dict[str, str]], **engine_kwargs: Any
+) -> Iterator[SQLAlchemyClient]:
+    url = url_from_config(url_or_config)
+    with handle_error(url):
+        engine = create_engine(url, **engine_kwargs)
+    try:
+        yield SQLAlchemyClient(engine)
+    finally:
+        engine.dispose()

dvc/dependency/__init__.py

@@ -4,6 +4,7 @@ from typing import Any, Dict, List, Mapping, Set
 from dvc.output import ARTIFACT_SCHEMA, DIR_FILES_SCHEMA, Output
 
 from .base import Dependency
+from .db import DB_SCHEMA, PARAM_DB, DbDependency, DbtDependency
 from .param import ParamsDependency
 from .repo import RepoDependency
 
@@ -14,20 +15,28 @@ from .repo import RepoDependency
 SCHEMA: Mapping[str, Any] = {
     **ARTIFACT_SCHEMA,
     **RepoDependency.REPO_SCHEMA,
+    **DB_SCHEMA,
     Output.PARAM_FILES: [DIR_FILES_SCHEMA],
     Output.PARAM_FS_CONFIG: dict,
 }
 
 
 def _get(stage, p, info, **kwargs):
-    if info and info.get(RepoDependency.PARAM_REPO):
-        repo = info.pop(RepoDependency.PARAM_REPO)
-        return RepoDependency(repo, stage, p, info)
+    d = info or {}
+    db = d.get(PARAM_DB, {})
+    params = d.pop(ParamsDependency.PARAM_PARAMS, None)
+    repo = d.pop(RepoDependency.PARAM_REPO, None)
 
-    if info and info.get(ParamsDependency.PARAM_PARAMS):
-        params = info.pop(ParamsDependency.PARAM_PARAMS)
+    if params:
         return ParamsDependency(stage, p, params)
+    if DbDependency.PARAM_QUERY in db:
+        return DbDependency(stage, info)
+    if db:
+        return DbtDependency(repo, stage, info)
 
+    assert p
+    if repo:
+        return RepoDependency(repo, stage, p, info)
     return Dependency(stage, p, info, **kwargs)
 
 
@@ -44,9 +53,11 @@ def loadd_from(stage, d_list):
     return ret
 
 
-def loads_from(stage, s_list, erepo=None, fs_config=None):
+def loads_from(stage, s_list, erepo=None, fs_config=None, db=None):
     assert isinstance(s_list, list)
     info = {RepoDependency.PARAM_REPO: erepo} if erepo else {}
+    if db:
+        info.update({"db": db})
     return [
         _get(
             stage,

dvc/dependency/base.py

@@ -33,7 +33,7 @@ class Dependency(Output):
         if self.fs.version_aware:
             old_fs_path = self.fs_path
             try:
-                self.fs_path = self.fs.path.version_path(self.fs_path, None)
+                self.fs_path = self.fs.version_path(self.fs_path, None)
                 if self.changed_meta():
                     return {str(self): "update available"}
             finally:
@@ -42,9 +42,9 @@ class Dependency(Output):
 
     def update(self, rev=None):
         if self.fs.version_aware:
-            self.fs_path = self.fs.path.version_path(self.fs_path, rev)
+            self.fs_path = self.fs.version_path(self.fs_path, rev)
             self.meta = self.get_meta()
-            self.fs_path = self.fs.path.version_path(self.fs_path, self.meta.version_id)
+            self.fs_path = self.fs.version_path(self.fs_path, self.meta.version_id)
 
     def download(self, to, jobs=None):
         fs_download(self.fs, self.fs_path, to.fs_path, jobs=jobs)
@@ -52,7 +52,7 @@ class Dependency(Output):
     def save(self):
         super().save()
         if self.fs.version_aware:
-            self.fs_path = self.fs.path.version_path(self.fs_path, self.meta.version_id)
+            self.fs_path = self.fs.version_path(self.fs_path, self.meta.version_id)
 
     def dumpd(self, **kwargs):
         if self.fs.version_aware:

dvc/dependency/db.py

@@ -0,0 +1,324 @@
+import os
+from contextlib import contextmanager, nullcontext
+from typing import TYPE_CHECKING, Any, Callable, Dict, Iterator, Optional, Union
+
+from funcy import compact, log_durations
+
+from dvc.exceptions import DvcException
+from dvc.log import logger
+from dvc.scm import SCM
+
+from .base import Dependency
+
+if TYPE_CHECKING:
+    from rich.status import Status
+
+    from dvc.output import Output
+    from dvc.stage import Stage
+
+logger = logger.getChild(__name__)
+
+
+PARAM_DB = "db"
+PARAM_PROFILE = "profile"
+PARAM_FILE_FORMAT = "file_format"
+
+
+def _get_dbt_config(config: Dict) -> Dict:
+    conf = config.get("feature", {})
+    pref = "dbt_"
+    return {k.lstrip(pref): v for k, v in conf.items() if k.startswith(pref)}
+
+
+@contextmanager
+def log_status(
+    msg, status: Optional["Status"] = None, log=logger.debug
+) -> Iterator["Status"]:
+    from funcy import log_durations
+
+    from dvc.ui import ui
+
+    with log_durations(log, msg), status or ui.status(msg) as st:
+        st.update(msg)
+        yield st
+
+
+@contextmanager
+def chdir(path):
+    wdir = os.getcwd()
+    os.chdir(path)
+    try:
+        yield
+    finally:
+        os.chdir(wdir)
+
+
+@contextmanager
+def download_progress(to: "Output") -> Iterator[Callable[[int], Any]]:
+    from dvc.ui import ui
+    from dvc.ui._rich_progress import DbDownloadProgress
+
+    with log_durations(logger.debug, f"Saving to {to}"), DbDownloadProgress(
+        console=ui.error_console,
+    ) as progress:
+        task = progress.add_task("Saving", total=None, output=to)
+        yield lambda n: progress.advance(task, advance=n)
+        progress.update(task, description="Saved", total=0)
+
+
+class AbstractDependency(Dependency):
+    """Dependency without workspace/fs/fs_path"""
+
+    def __init__(self, stage: "Stage", info: Dict[str, Any], *args, **kwargs):
+        self.repo = stage.repo
+        self.stage = stage
+        self.fs = None
+        self.fs_path = None
+        self.def_path = None  # type: ignore[assignment]
+        self.info = info or {}
+
+
+class DbDependency(AbstractDependency):
+    PARAM_QUERY = "query"
+    PARAM_CONNECTION = "connection"
+    QUERY_SCHEMA = {PARAM_QUERY: str, PARAM_CONNECTION: str}
+
+    def __init__(self, stage: "Stage", info, *args, **kwargs):
+        super().__init__(stage, info, *args, **kwargs)
+        self.target = None
+
+    def __repr__(self):
+        return "{}:{}".format(
+            self.__class__.__name__, "".join(f"{k}=={v}" for k, v in self.info.items())
+        )
+
+    def __str__(self):
+        from dvc.utils.humanize import truncate_text
+
+        db_info = self.info.get(PARAM_DB, {})
+        query = db_info.get(self.PARAM_QUERY, "[query]")
+        return truncate_text(query, 50)
+
+    def workspace_status(self):
+        return False  # no workspace to check
+
+    def status(self):
+        return self.workspace_status()
+
+    def save(self):
+        """nothing to save."""
+
+    def dumpd(self, **kwargs):
+        db_info = compact(self.info.get(PARAM_DB, {}))
+        return {PARAM_DB: db_info} if db_info else {}
+
+    def update(self, rev=None):
+        """nothing to update."""
+
+    def download(
+        self,
+        to: "Output",
+        jobs: Optional[int] = None,  # noqa: ARG002
+        file_format: Optional[str] = None,
+        **kwargs: Any,
+    ) -> None:
+        from dvc.database import export, get_client
+
+        db_info = self.info.get(PARAM_DB, {})
+        query = db_info.get(self.PARAM_QUERY)
+        if not query:
+            raise DvcException("Cannot download: no query specified")
+
+        dbt_config = _get_dbt_config(self.repo.config)
+        profile = db_info.get(PARAM_PROFILE) or dbt_config.get(PARAM_PROFILE)
+        target = self.target or dbt_config.get("target")
+
+        connection = db_info.get(self.PARAM_CONNECTION)
+        db_config = self.repo.config.get("db", {})
+        config = db_config.get(connection)
+        if connection and not config:
+            raise DvcException(f"connection {connection} not found in config")
+
+        project_dir = self.repo.root_dir
+        with get_client(
+            config, project_dir=project_dir, profile=profile, target=target
+        ) as db:
+            logger.debug("using adapter: %s", db)
+            with log_status("Testing connection") as status:
+                db.test_connection(onerror=status.stop)
+
+            file_format = file_format or db_info.get(PARAM_FILE_FORMAT, "csv")
+            assert file_format
+            with log_status("Executing query") as status, db.query(query) as serializer:
+                status.stop()
+                logger.debug("using serializer: %s", serializer)
+                with download_progress(to) as progress:
+                    return export(
+                        serializer, to.fs_path, format=file_format, progress=progress
+                    )
+
+
+class DbtDependency(AbstractDependency):
+    PARAM_MODEL = "model"
+    PARAM_VERSION = "version"
+    PARAM_PROJECT_DIR = "project_dir"
+    DBT_SCHEMA = {
+        PARAM_MODEL: str,
+        PARAM_VERSION: str,
+        PARAM_PROJECT_DIR: str,
+    }
+
+    def __init__(self, def_repo: Dict[str, Any], stage: "Stage", info, *args, **kwargs):
+        self.def_repo = def_repo or {}
+        self.target = None
+        super().__init__(stage, info, *args, **kwargs)
+
+    def __repr__(self):
+        return "{}:{}".format(
+            self.__class__.__name__,
+            "".join(f"{k}=={v}" for k, v in {**self.def_repo, **self.info}.items()),
+        )
+
+    def __str__(self):
+        from .repo import RepoDependency
+
+        repo = self.def_repo.get(RepoDependency.PARAM_URL)
+        rev = self.def_repo.get(RepoDependency.PARAM_REV)
+
+        db_info = self.info.get(PARAM_DB, {})
+        db = db_info.get(self.PARAM_MODEL, "")
+        project_dir = db_info.get(self.PARAM_PROJECT_DIR, "")
+        repo_info = ""
+        if repo:
+            repo_info += repo
+        if rev:
+            repo_info += f"@{rev}"
+        if project_dir:
+            repo_info += f":/{project_dir}"
+        return db + (f"({repo_info})" if repo_info else "")
+
+    @property
+    def locked_rev(self):
+        from .repo import RepoDependency
+
+        return self.def_repo.get(RepoDependency.PARAM_REV_LOCK)
+
+    @property
+    def rev(self):
+        from .repo import RepoDependency
+
+        return self.def_repo.get(RepoDependency.PARAM_REV)
+
+    def workspace_status(self):
+        if not self.def_repo:
+            return
+
+        current = self._get_clone(self.locked_rev or self.rev).get_rev()
+        updated = self._get_clone(self.rev).get_rev()
+        if current != updated:
+            return {str(self): "update available"}
+        return {}
+
+    def status(self):
+        return self.workspace_status()
+
+    def save(self):
+        from .repo import RepoDependency
+
+        if not self.def_repo:
+            return
+
+        rev = self._get_clone(self.locked_rev or self.rev).get_rev()
+        if self.def_repo.get(RepoDependency.PARAM_REV_LOCK) is None:
+            self.def_repo[RepoDependency.PARAM_REV_LOCK] = rev
+
+    def dumpd(self, **kwargs) -> Dict[str, Union[str, Dict[str, str]]]:
+        from .repo import RepoDependency
+
+        def_repo = {}
+        if self.def_repo:
+            def_repo = RepoDependency._dump_def_repo(self.def_repo)
+
+        db_info = compact(self.info.get(PARAM_DB, {}))
+        return compact({RepoDependency.PARAM_REPO: def_repo, PARAM_DB: db_info})
+
+    def _get_clone(self, rev):
+        from dvc.repo.open_repo import _cached_clone
+
+        from .repo import RepoDependency
+
+        url = self.def_repo.get(RepoDependency.PARAM_URL)
+        repo_root = self.repo.root_dir if self.repo else os.getcwd()
+        return SCM(_cached_clone(url, rev) if url else repo_root)
+
+    def update(self, rev=None):
+        from .repo import RepoDependency
+
+        if not self.def_repo:
+            return
+
+        if rev:
+            self.def_repo[RepoDependency.PARAM_REV] = rev
+        else:
+            rev = self.rev
+        self.def_repo[RepoDependency.PARAM_REV_LOCK] = self._get_clone(rev).get_rev()
+
+    def download(
+        self,
+        to: "Output",
+        jobs: Optional[int] = None,  # noqa: ARG002
+        file_format: Optional[str] = None,
+    ) -> None:
+        from .repo import RepoDependency
+
+        project_dir = self.info.get(PARAM_DB, {}).get(self.PARAM_PROJECT_DIR, "")
+        if self.def_repo:
+            ctx = repo = self._get_clone(self.locked_rev or self.rev)
+            self.def_repo[RepoDependency.PARAM_REV_LOCK] = repo.get_rev()
+            root = wdir = repo.root_dir
+        else:
+            ctx = nullcontext()
+            root = self.repo.root_dir
+            wdir = self.stage.wdir
+
+        project_path = os.path.join(wdir, project_dir) if project_dir else root
+        with ctx, chdir(project_path):
+            self._download_db(to, file_format=file_format)
+
+    def _download_db(
+        self,
+        to: "Output",
+        version: Optional[int] = None,
+        file_format: Optional[str] = None,
+    ) -> None:
+        from dvc.database import export, get_model
+
+        db_info = self.info.get(PARAM_DB, {})
+        model = db_info.get(self.PARAM_MODEL)
+        version = version or db_info.get(self.PARAM_VERSION)
+        file_format = file_format or db_info.get(PARAM_FILE_FORMAT) or "csv"
+        assert file_format
+        if not model:
+            raise DvcException("Cannot download, no model specified")
+
+        dbt_config = _get_dbt_config(self.repo.config)
+        profile = db_info.get(PARAM_PROFILE) or dbt_config.get(PARAM_PROFILE)
+        target = self.target or db_info.get("target") or dbt_config.get("target")
+
+        with log_status("Downloading model"):
+            serializer = get_model(
+                model, version=version, profile=profile, target=target
+            )
+        # NOTE: we keep everything in memory, and then export it out later.
+        with download_progress(to) as progress:
+            export(serializer, to.fs_path, format=file_format, progress=progress)
+
+
+DB_SCHEMA = {
+    PARAM_DB: {
+        PARAM_PROFILE: str,
+        PARAM_FILE_FORMAT: str,
+        **DbDependency.QUERY_SCHEMA,
+        **DbtDependency.DBT_SCHEMA,
+    },
+}

dvc/dependency/param.py

@@ -54,12 +54,15 @@ def read_param_file(
                 continue
         return ret
 
-    from dpath import merge
+    from copy import deepcopy
 
-    for key_path in key_paths:
+    from dpath import merge
+    from funcy import distinct
+
+    for key_path in distinct(key_paths):
         merge(
             ret,
-            dpath.search(config, key_path, separator="."),
+            deepcopy(dpath.search(config, key_path, separator=".")),
             separator=".",
         )
     return ret

dvc/dependency/repo.py

@@ -64,28 +64,31 @@ class RepoDependency(Dependency):
         if self.def_repo.get(self.PARAM_REV_LOCK) is None:
             self.def_repo[self.PARAM_REV_LOCK] = rev
 
-    def dumpd(self, **kwargs) -> Dict[str, Union[str, Dict[str, str]]]:
-        repo = {self.PARAM_URL: self.def_repo[self.PARAM_URL]}
+    @classmethod
+    def _dump_def_repo(cls, def_repo) -> Dict[str, str]:
+        repo = {cls.PARAM_URL: def_repo[cls.PARAM_URL]}
 
-        rev = self.def_repo.get(self.PARAM_REV)
+        rev = def_repo.get(cls.PARAM_REV)
         if rev:
-            repo[self.PARAM_REV] = self.def_repo[self.PARAM_REV]
+            repo[cls.PARAM_REV] = def_repo[cls.PARAM_REV]
 
-        rev_lock = self.def_repo.get(self.PARAM_REV_LOCK)
+        rev_lock = def_repo.get(cls.PARAM_REV_LOCK)
         if rev_lock:
-            repo[self.PARAM_REV_LOCK] = rev_lock
+            repo[cls.PARAM_REV_LOCK] = rev_lock
 
-        config = self.def_repo.get(self.PARAM_CONFIG)
+        config = def_repo.get(cls.PARAM_CONFIG)
         if config:
-            repo[self.PARAM_CONFIG] = config
+            repo[cls.PARAM_CONFIG] = config
 
-        remote = self.def_repo.get(self.PARAM_REMOTE)
+        remote = def_repo.get(cls.PARAM_REMOTE)
         if remote:
-            repo[self.PARAM_REMOTE] = remote
+            repo[cls.PARAM_REMOTE] = remote
+        return repo
 
+    def dumpd(self, **kwargs) -> Dict[str, Union[str, Dict[str, str]]]:
         return {
             self.PARAM_PATH: self.def_path,
-            self.PARAM_REPO: repo,
+            self.PARAM_REPO: self._dump_def_repo(self.def_repo),
         }
 
     def update(self, rev: Optional[str] = None):

dvc/dirs.py

@@ -15,7 +15,7 @@ def system_config_dir():
 
 
 def global_config_dir():
-    return os.getenv(env.DVC_SYSTEM_CONFIG_DIR) or platformdirs.user_config_dir(
+    return os.getenv(env.DVC_GLOBAL_CONFIG_DIR) or platformdirs.user_config_dir(
         APPNAME, APPAUTHOR
     )
 

dvc/dvcfile.py

@@ -294,7 +294,7 @@ class ProjectFile(FileMixin):
     def resolver(self) -> "DataResolver":
         from .parsing import DataResolver
 
-        wdir = self.repo.fs.path.parent(self.path)
+        wdir = self.repo.fs.parent(self.path)
         return DataResolver(self.repo, wdir, self.contents)
 
     @cached_property

dvc/fs/__init__.py

@@ -5,6 +5,8 @@ from dvc_http import HTTPFileSystem, HTTPSFileSystem  # noqa: F401
 
 from dvc.config import ConfigError as RepoConfigError
 from dvc.config_schema import SCHEMA, Invalid
+
+# pylint: disable=unused-import
 from dvc_objects.fs import (  # noqa: F401
     LocalFileSystem,
     MemoryFileSystem,
@@ -23,11 +25,10 @@ from dvc_objects.fs.errors import (  # noqa: F401
     ConfigError,
     RemoteMissingDepsError,
 )
-from dvc_objects.fs.path import Path  # noqa: F401
 
-from .callbacks import Callback
+from .callbacks import Callback  # noqa: F401
 from .data import DataFileSystem  # noqa: F401
-from .dvc import DVCFileSystem  # noqa: F401
+from .dvc import DVCFileSystem
 from .git import GitFileSystem  # noqa: F401
 
 known_implementations.update(
@@ -47,29 +48,32 @@ known_implementations.update(
 def download(
     fs: "FileSystem", fs_path: str, to: str, jobs: Optional[int] = None
 ) -> int:
-    with Callback.as_tqdm_callback(
-        desc=f"Downloading {fs.path.name(fs_path)}",
+    from dvc.scm import lfs_prefetch
+
+    from .callbacks import TqdmCallback
+
+    with TqdmCallback(
+        desc=f"Downloading {fs.name(fs_path)}",
         unit="files",
     ) as cb:
         # NOTE: We use dvc-objects generic.copy over fs.get since it makes file
         # download atomic and avoids fsspec glob/regex path expansion.
         if fs.isdir(fs_path):
             from_infos = [
-                path
-                for path in fs.find(fs_path)
-                if not path.endswith(fs.path.flavour.sep)
+                path for path in fs.find(fs_path) if not path.endswith(fs.flavour.sep)
             ]
             if not from_infos:
                 localfs.makedirs(to, exist_ok=True)
                 return 0
             to_infos = [
-                localfs.path.join(to, *fs.path.relparts(info, fs_path))
-                for info in from_infos
+                localfs.join(to, *fs.relparts(info, fs_path)) for info in from_infos
             ]
         else:
             from_infos = [fs_path]
             to_infos = [to]
 
+        if isinstance(fs, DVCFileSystem):
+            lfs_prefetch(fs, from_infos)
         cb.set_size(len(from_infos))
         jobs = jobs or fs.jobs
         generic.copy(

dvc/fs/data.py

@@ -29,6 +29,9 @@ class DataFileSystem(FileSystem):
 
         return _DataFileSystem(**self.fs_args)
 
+    def getcwd(self):
+        return self.fs.getcwd()
+
     def isdvc(self, path, **kwargs):
         return self.fs.isdvc(path, **kwargs)
 

dvc/fs/dvc.py

@@ -4,6 +4,7 @@ import ntpath
 import os
 import posixpath
 import threading
+from collections import deque
 from contextlib import ExitStack, suppress
 from typing import TYPE_CHECKING, Any, Callable, Dict, Optional, Tuple, Type, Union
 
@@ -12,7 +13,6 @@ from funcy import wrap_with
 
 from dvc.log import logger
 from dvc_objects.fs.base import FileSystem
-from dvc_objects.fs.path import Path
 
 from .data import DataFileSystem
 
@@ -60,6 +60,7 @@ def _merge_info(repo, key, fs_info, dvc_info):
     if fs_info:
         ret["type"] = fs_info["type"]
         ret["size"] = fs_info["size"]
+        ret["fs_info"] = fs_info
         isexec = False
         if fs_info["type"] == "file":
             isexec = utils.is_exec(fs_info["mode"])
@@ -69,7 +70,7 @@ def _merge_info(repo, key, fs_info, dvc_info):
 
 
 def _get_dvc_path(dvc_fs, subkey):
-    return dvc_fs.path.join(*subkey) if subkey else ""
+    return dvc_fs.join(*subkey) if subkey else ""
 
 
 class _DVCFileSystem(AbstractFileSystem):
@@ -129,55 +130,88 @@ class _DVCFileSystem(AbstractFileSystem):
             ...    rev="main",
             ... )
         """
-        from pygtrie import Trie
-
         super().__init__()
+        self._repo = repo
+        self._repo_factory = repo_factory
+        self._traverse_subrepos = subrepos
         self._repo_stack = ExitStack()
-        if repo is None:
-            url = url if url is not None else fo
-            repo = self._make_repo(
-                url=url,
-                rev=rev,
-                subrepos=subrepos,
-                config=config,
-                remote=remote,
-                remote_config=remote_config,
-            )
-            assert repo is not None
+        self._repo_kwargs = {
+            "url": url if url is not None else fo,
+            "rev": rev,
+            "subrepos": subrepos,
+            "config": config,
+            "remote": remote,
+            "remote_config": remote_config,
+        }
 
-            repo_factory = repo._fs_conf["repo_factory"]
-            self._repo_stack.enter_context(repo)
+    def getcwd(self):
+        relparts: Tuple[str, ...] = ()
+        assert self.repo is not None
+        if self.repo.fs.isin(self.repo.fs.getcwd(), self.repo.root_dir):
+            relparts = self.repo.fs.relparts(self.repo.fs.getcwd(), self.repo.root_dir)
+        return self.root_marker + self.sep.join(relparts)
 
-        if not repo_factory:
+    @classmethod
+    def join(cls, *parts: str) -> str:
+        return posixpath.join(*parts)
+
+    @classmethod
+    def parts(cls, path: str) -> Tuple[str, ...]:
+        ret = []
+        while True:
+            path, part = posixpath.split(path)
+
+            if part:
+                ret.append(part)
+                continue
+
+            if path:
+                ret.append(path)
+
+            break
+
+        ret.reverse()
+
+        return tuple(ret)
+
+    def normpath(self, path: str) -> str:
+        return posixpath.normpath(path)
+
+    def abspath(self, path: str) -> str:
+        if not posixpath.isabs(path):
+            path = self.join(self.getcwd(), path)
+        return self.normpath(path)
+
+    def relpath(self, path: str, start: Optional[str] = None) -> str:
+        if start is None:
+            start = "."
+        return posixpath.relpath(self.abspath(path), start=self.abspath(start))
+
+    def relparts(self, path: str, start: Optional[str] = None) -> Tuple[str, ...]:
+        return self.parts(self.relpath(path, start=start))
+
+    @functools.cached_property
+    def repo(self):
+        if self._repo:
+            return self._repo
+
+        repo = self._make_repo(**self._repo_kwargs)
+
+        self._repo_stack.enter_context(repo)
+        self._repo = repo
+        return repo
+
+    @functools.cached_property
+    def repo_factory(self):
+        if self._repo_factory:
+            return self._repo_factory
+
+        if self._repo:
             from dvc.repo import Repo
 
-            self.repo_factory: RepoFactory = Repo
-        else:
-            self.repo_factory = repo_factory
+            return Repo
 
-        def _getcwd():
-            relparts: Tuple[str, ...] = ()
-            assert repo is not None
-            if repo.fs.path.isin(repo.fs.path.getcwd(), repo.root_dir):
-                relparts = repo.fs.path.relparts(repo.fs.path.getcwd(), repo.root_dir)
-            return self.root_marker + self.sep.join(relparts)
-
-        self.path = Path(self.sep, getcwd=_getcwd)
-        self.repo = repo
-        self.hash_jobs = repo.fs.hash_jobs
-        self._traverse_subrepos = subrepos
-
-        self._subrepos_trie = Trie()
-        """Keeps track of each and every path with the corresponding repo."""
-
-        key = self._get_key(self.repo.root_dir)
-        self._subrepos_trie[key] = repo
-
-        self._datafss = {}
-        """Keep a datafs instance of each repo."""
-
-        if hasattr(repo, "dvc_dir"):
-            self._datafss[key] = DataFileSystem(index=repo.index.data["repo"])
+        return self.repo._fs_conf["repo_factory"]
 
     @functools.cached_property
     def fsid(self) -> str:
@@ -192,20 +226,43 @@ class _DVCFileSystem(AbstractFileSystem):
 
     def _get_key(self, path: "StrPath") -> Key:
         path = os.fspath(path)
-        parts = self.repo.fs.path.relparts(path, self.repo.root_dir)
+        parts = self.repo.fs.relparts(path, self.repo.root_dir)
         if parts == (os.curdir,):
             return ()
         return parts
 
+    @functools.cached_property
+    def _subrepos_trie(self):
+        """Keeps track of each and every path with the corresponding repo."""
+
+        from pygtrie import Trie
+
+        trie = Trie()
+        key = self._get_key(self.repo.root_dir)
+        trie[key] = self.repo
+        return trie
+
     def _get_key_from_relative(self, path) -> Key:
         path = self._strip_protocol(path)
-        parts = self.path.relparts(path, self.root_marker)
+        parts = self.relparts(path, self.root_marker)
         if parts and parts[0] == os.curdir:
             return parts[1:]
         return parts
 
     def _from_key(self, parts: Key) -> str:
-        return self.repo.fs.path.join(self.repo.root_dir, *parts)
+        return self.repo.fs.join(self.repo.root_dir, *parts)
+
+    @functools.cached_property
+    def _datafss(self):
+        """Keep a datafs instance of each repo."""
+
+        datafss = {}
+
+        if hasattr(self.repo, "dvc_dir"):
+            key = self._get_key(self.repo.root_dir)
+            datafss[key] = DataFileSystem(index=self.repo.index.data["repo"])
+
+        return datafss
 
     @property
     def repo_url(self):
@@ -259,7 +316,7 @@ class _DVCFileSystem(AbstractFileSystem):
 
         from dvc.repo import Repo
 
-        repo_path = self.repo.fs.path.join(dir_path, Repo.DVC_DIR)
+        repo_path = self.repo.fs.join(dir_path, Repo.DVC_DIR)
         return self.repo.fs.isdir(repo_path)
 
     def _get_subrepo_info(
@@ -313,7 +370,7 @@ class _DVCFileSystem(AbstractFileSystem):
             dvc_path = _get_dvc_path(dvc_fs, subkey)
             with suppress(FileNotFoundError, NotADirectoryError):
                 for info in dvc_fs.ls(dvc_path, detail=True):
-                    dvc_infos[dvc_fs.path.name(info["name"])] = info
+                    dvc_infos[dvc_fs.name(info["name"])] = info
                 dvc_exists = True
 
         fs_exists = False
@@ -326,7 +383,7 @@ class _DVCFileSystem(AbstractFileSystem):
                 for info in repo.dvcignore.ls(
                     fs, fs_path, detail=True, ignore_subrepos=ignore_subrepos
                 ):
-                    fs_infos[fs.path.name(info["name"])] = info
+                    fs_infos[fs.name(info["name"])] = info
                 fs_exists = True
             except (FileNotFoundError, NotADirectoryError):
                 pass
@@ -345,7 +402,7 @@ class _DVCFileSystem(AbstractFileSystem):
             if not dvcfiles and _is_dvc_file(name):
                 continue
 
-            entry_path = self.path.join(path, name)
+            entry_path = self.join(path, name)
             info = _merge_info(
                 repo, (*subkey, name), fs_infos.get(name), dvc_infos.get(name)
             )
@@ -393,7 +450,7 @@ class _DVCFileSystem(AbstractFileSystem):
         # NOTE: if some parent in fs_path turns out to be a file, it means
         # that the whole repofs branch doesn't exist.
         if dvc_info and not fs_info:
-            for parent in fs.path.parents(fs_path):
+            for parent in fs.parents(fs_path):
                 try:
                     if fs.info(parent)["type"] != "directory":
                         dvc_info = None
@@ -421,6 +478,45 @@ class _DVCFileSystem(AbstractFileSystem):
         dvc_path = _get_dvc_path(dvc_fs, subkey)
         return dvc_fs.get_file(dvc_path, lpath, **kwargs)
 
+    def du(self, path, total=True, maxdepth=None, withdirs=False, **kwargs):
+        if maxdepth is not None:
+            raise NotImplementedError
+
+        sizes = {}
+        dus = {}
+        todo = deque([self.info(path)])
+        while todo:
+            info = todo.popleft()
+            isdir = info["type"] == "directory"
+            size = info["size"] or 0
+            name = info["name"]
+
+            if not isdir:
+                sizes[name] = size
+                continue
+
+            dvc_info = info.get("dvc_info") or {}
+            fs_info = info.get("fs_info")
+            entry = dvc_info.get("entry")
+            if (
+                dvc_info
+                and not fs_info
+                and entry is not None
+                and entry.size is not None
+            ):
+                dus[name] = entry.size
+                continue
+
+            if withdirs:
+                sizes[name] = size
+
+            todo.extend(self.ls(info["name"], detail=True))
+
+        if total:
+            return sum(sizes.values()) + sum(dus.values())
+
+        return sizes
+
     def close(self):
         self._repo_stack.close()
 
@@ -436,6 +532,9 @@ class DVCFileSystem(FileSystem):
     def fs(self) -> "_DVCFileSystem":
         return _DVCFileSystem(**self.fs_args)
 
+    def getcwd(self):
+        return self.fs.getcwd()
+
     @property
     def fsid(self) -> str:
         return self.fs.fsid
@@ -443,10 +542,6 @@ class DVCFileSystem(FileSystem):
     def isdvc(self, path, **kwargs) -> bool:
         return self.fs.isdvc(path, **kwargs)
 
-    @property
-    def path(self) -> Path:
-        return self.fs.path
-
     @property
     def repo(self) -> "Repo":
         return self.fs.repo

dvc/fs/git.py

@@ -46,9 +46,11 @@ class GitFileSystem(FileSystem):
 
         return FsspecGitFileSystem(**self.fs_args)
 
-    @functools.cached_property
-    def path(self):
-        return self.fs.path
+    def getcwd(self):
+        return self.fs.getcwd()
+
+    def chdir(self, path):
+        self.fs.chdir(path)
 
     @property
     def rev(self) -> str:

dvc/ignore.py

@@ -51,8 +51,8 @@ class DvcIgnorePatterns(DvcIgnore):
 
     @classmethod
     def from_file(cls, path, fs, name):
-        assert fs.path.isabs(path)
-        dirname = fs.path.normpath(fs.path.dirname(path))
+        assert fs.isabs(path)
+        dirname = fs.normpath(fs.dirname(path))
         with fs.open(path, encoding="utf-8") as fobj:
             path_spec_lines = [
                 PatternInfo(line, f"{name}:{line_no + 1}:{line}")
@@ -194,7 +194,7 @@ class DvcIgnoreFilter:
         )
 
     def _get_key(self, path):
-        parts = self.fs.path.relparts(path, self.root_dir)
+        parts = self.fs.relparts(path, self.root_dir)
         if parts == (os.curdir,):
             return ()
         return parts
@@ -204,13 +204,13 @@ class DvcIgnoreFilter:
         old_pattern = trie.longest_prefix(key).value
         matches = old_pattern.matches(dirname, DvcIgnore.DVCIGNORE_FILE, False)
 
-        path = self.fs.path.join(dirname, DvcIgnore.DVCIGNORE_FILE)
+        path = self.fs.join(dirname, DvcIgnore.DVCIGNORE_FILE)
         if not matches and self.fs.exists(path):
-            name = self.fs.path.relpath(path, self.root_dir)
+            name = self.fs.relpath(path, self.root_dir)
             new_pattern = DvcIgnorePatterns.from_file(path, self.fs, name)
             if old_pattern:
                 plist, prefix = merge_patterns(
-                    self.fs.path.flavour,
+                    self.fs.flavour,
                     old_pattern.pattern_list,
                     old_pattern.dirname,
                     new_pattern.pattern_list,
@@ -239,7 +239,7 @@ class DvcIgnoreFilter:
                     dnames = []
 
             for dname in dnames:
-                self._update_sub_repo(self.fs.path.join(dirname, dname), ignore_trie)
+                self._update_sub_repo(self.fs.join(dirname, dname), ignore_trie)
 
     def _update_sub_repo(self, path, ignore_trie: Trie):
         from dvc.repo import Repo
@@ -247,18 +247,18 @@ class DvcIgnoreFilter:
         if path == self.root_dir:
             return
 
-        dvc_dir = self.fs.path.join(path, Repo.DVC_DIR)
+        dvc_dir = self.fs.join(path, Repo.DVC_DIR)
         if not self.fs.exists(dvc_dir):
             return
 
-        root, dname = self.fs.path.split(path)
+        root, dname = self.fs.split(path)
         key = self._get_key(root)
         pattern_info = PatternInfo(f"/{dname}/", f"in sub_repo:{dname}")
         new_pattern = DvcIgnorePatterns([pattern_info], root, self.fs.sep)
         old_pattern = ignore_trie.longest_prefix(key).value
         if old_pattern:
             plist, prefix = merge_patterns(
-                self.fs.path.flavour,
+                self.fs.flavour,
                 old_pattern.pattern_list,
                 old_pattern.dirname,
                 new_pattern.pattern_list,
@@ -269,7 +269,7 @@ class DvcIgnoreFilter:
             ignore_trie[key] = new_pattern
 
     def __call__(self, root, dirs, files, ignore_subrepos=True):
-        abs_root = self.fs.path.abspath(root)
+        abs_root = self.fs.abspath(root)
         ignore_pattern = self._get_trie_pattern(
             abs_root, dnames=dirs, ignore_subrepos=ignore_subrepos
         )
@@ -283,7 +283,7 @@ class DvcIgnoreFilter:
         nondirs = []
 
         for entry in fs.ls(path, detail=True, **kwargs):
-            name = fs.path.name(entry["name"])
+            name = fs.name(entry["name"])
             fs_dict[name] = entry
             if entry["type"] == "directory":
                 dirs.append(name)
@@ -338,7 +338,7 @@ class DvcIgnoreFilter:
         else:
             ignores_trie = self._ignores_trie_subrepos
 
-        if not self.fs.path.isin_or_eq(dirname, self.root_dir):
+        if not self.fs.isin_or_eq(dirname, self.root_dir):
             # outside of the repo
             return None
 
@@ -349,12 +349,12 @@ class DvcIgnoreFilter:
             return ignore_pattern
 
         prefix_key = ignores_trie.longest_prefix(key).key or ()
-        prefix = self.fs.path.join(self.root_dir, *prefix_key)
+        prefix = self.fs.join(self.root_dir, *prefix_key)
 
         dirs = list(
             takewhile(
                 lambda path: path != prefix,
-                (parent for parent in localfs.path.parents(dirname)),
+                (parent for parent in localfs.parents(dirname)),
             )
         )
         dirs.reverse()
@@ -370,7 +370,7 @@ class DvcIgnoreFilter:
     ):
         if self._outside_repo(path):
             return False
-        dirname, basename = self.fs.path.split(self.fs.path.normpath(path))
+        dirname, basename = self.fs.split(self.fs.normpath(path))
         ignore_pattern = self._get_trie_pattern(dirname, None, ignore_subrepos)
         if ignore_pattern:
             return ignore_pattern.matches(dirname, basename, is_dir)
@@ -378,7 +378,7 @@ class DvcIgnoreFilter:
 
     def is_ignored_dir(self, path: str, ignore_subrepos: bool = True) -> bool:
         # only used in LocalFileSystem
-        path = self.fs.path.abspath(path)
+        path = self.fs.abspath(path)
         if path == self.root_dir:
             return False
 
@@ -386,18 +386,18 @@ class DvcIgnoreFilter:
 
     def is_ignored_file(self, path: str, ignore_subrepos: bool = True) -> bool:
         # only used in LocalFileSystem
-        path = self.fs.path.abspath(path)
+        path = self.fs.abspath(path)
         return self._is_ignored(path, False, ignore_subrepos=ignore_subrepos)
 
     def _outside_repo(self, path):
-        return not self.fs.path.isin_or_eq(path, self.root_dir)
+        return not self.fs.isin_or_eq(path, self.root_dir)
 
     def check_ignore(self, target):
         # NOTE: can only be used in `dvc check-ignore`, see
         # https://github.com/iterative/dvc/issues/5046
-        full_target = self.fs.path.abspath(target)
+        full_target = self.fs.abspath(target)
         if not self._outside_repo(full_target):
-            dirname, basename = self.fs.path.split(self.fs.path.normpath(full_target))
+            dirname, basename = self.fs.split(self.fs.normpath(full_target))
             pattern = self._get_trie_pattern(dirname)
             if pattern:
                 matches = pattern.matches(

dvc/output.py

@@ -378,7 +378,7 @@ class Output:
         if (
             self.repo
             and self.fs.protocol == "local"
-            and not self.fs.path.isabs(self.def_path)
+            and not self.fs.isabs(self.def_path)
         ):
             self.fs = self.repo.fs
 
@@ -411,7 +411,7 @@ class Output:
         self.remote = remote
 
         if self.fs.version_aware:
-            _, version_id = self.fs.path.coalesce_version(
+            _, version_id = self.fs.coalesce_version(
                 self.def_path, self.meta.version_id
             )
             self.meta.version_id = version_id
@@ -465,7 +465,7 @@ class Output:
             parsed.scheme != "remote"
             and self.stage
             and self.stage.repo.fs == fs
-            and not fs.path.isabs(fs_path)
+            and not fs.isabs(fs_path)
         ):
             # NOTE: we can path either from command line or .dvc file,
             # so we should expect both posix and windows style paths.
@@ -473,9 +473,9 @@ class Output:
             #
             # FIXME: if we have Windows path containing / or posix one with \
             # then we have #2059 bug and can't really handle that.
-            fs_path = fs.path.join(self.stage.wdir, fs_path)
+            fs_path = fs.join(self.stage.wdir, fs_path)
 
-        return fs.path.abspath(fs.path.normpath(fs_path))
+        return fs.abspath(fs.normpath(fs_path))
 
     def __repr__(self):
         return f"{type(self).__name__}: {self.def_path!r}"
@@ -491,14 +491,14 @@ class Output:
         ):
             return str(self.def_path)
 
-        if not self.fs.path.isin(self.fs_path, self.repo.root_dir):
+        if not self.fs.isin(self.fs_path, self.repo.root_dir):
             return self.fs_path
 
-        cur_dir = self.fs.path.getcwd()
-        if self.fs.path.isin(cur_dir, self.repo.root_dir):
-            return self.fs.path.relpath(self.fs_path, cur_dir)
+        cur_dir = self.fs.getcwd()
+        if self.fs.isin(cur_dir, self.repo.root_dir):
+            return self.fs.relpath(self.fs_path, cur_dir)
 
-        return self.fs.path.relpath(self.fs_path, self.repo.root_dir)
+        return self.fs.relpath(self.fs_path, self.repo.root_dir)
 
     def clear(self):
         self.hash_info = HashInfo.from_dict({})
@@ -515,10 +515,10 @@ class Output:
         if urlparse(self.def_path).scheme == "remote":
             return False
 
-        if self.fs.path.isabs(self.def_path):
+        if self.fs.isabs(self.def_path):
             return False
 
-        return self.repo and self.fs.path.isin(
+        return self.repo and self.fs.isin(
             self.fs_path,
             self.repo.root_dir,
         )
@@ -607,11 +607,11 @@ class Output:
     def index_key(self) -> Tuple[str, "DataIndexKey"]:
         if self.is_in_repo:
             workspace = "repo"
-            key = self.repo.fs.path.relparts(self.fs_path, self.repo.root_dir)
+            key = self.repo.fs.relparts(self.fs_path, self.repo.root_dir)
         else:
             workspace = self.fs.protocol
-            no_drive = self.fs.path.flavour.splitdrive(self.fs_path)[1]
-            key = self.fs.path.parts(no_drive)[1:]
+            no_drive = self.fs.flavour.splitdrive(self.fs_path)[1]
+            key = self.fs.parts(no_drive)[1:]
         return workspace, key
 
     def changed_checksum(self):
@@ -786,7 +786,7 @@ class Output:
                     self.hash_name,
                     ignore=self.dvcignore,
                 )
-                with Callback.as_tqdm_callback(
+                with TqdmCallback(
                     desc=f"Committing {self} to cache",
                     unit="file",
                 ) as cb:
@@ -799,7 +799,7 @@ class Output:
                         callback=cb,
                     )
             if relink:
-                rel = self.fs.path.relpath(filter_info or self.fs_path)
+                rel = self.fs.relpath(filter_info or self.fs_path)
                 with CheckoutCallback(desc=f"Checking out {rel}", unit="files") as cb:
                     self._checkout(
                         filter_info or self.fs_path,
@@ -814,7 +814,7 @@ class Output:
                 self.set_exec()
 
     def _commit_granular_dir(self, filter_info, hardlink) -> Optional["HashFile"]:
-        prefix = self.fs.path.parts(self.fs.path.relpath(filter_info, self.fs_path))
+        prefix = self.fs.parts(self.fs.relpath(filter_info, self.fs_path))
         staging, _, obj = self._build(
             self.cache,
             self.fs_path,
@@ -826,7 +826,7 @@ class Output:
         save_obj = obj.filter(prefix)
         assert isinstance(save_obj, Tree)
         checkout_obj = save_obj.get_obj(self.cache, prefix)
-        with Callback.as_tqdm_callback(
+        with TqdmCallback(
             desc=f"Committing {self} to cache",
             unit="file",
         ) as cb:
@@ -862,7 +862,7 @@ class Output:
             ret.update(split_file_meta_from_cloud(meta_d))
 
         if self.is_in_repo:
-            path = self.fs.path.as_posix(relpath(self.fs_path, self.stage.wdir))
+            path = self.fs.as_posix(relpath(self.fs_path, self.stage.wdir))
         else:
             path = self.def_path
 
@@ -949,7 +949,7 @@ class Output:
             return None
 
         assert obj
-        fs_path = self.fs.path
+        fs_path = self.fs
         if filter_info and filter_info != self.fs_path:
             prefix = fs_path.relparts(filter_info, self.fs_path)
             assert isinstance(obj, Tree)
@@ -1059,7 +1059,7 @@ class Output:
             upload=upload,
             no_progress_bar=no_progress_bar,
         )
-        with Callback.as_tqdm_callback(
+        with TqdmCallback(
             desc=f"Transferring to {odb.fs.unstrip_protocol(odb.path)}",
             unit="file",
         ) as cb:
@@ -1092,7 +1092,10 @@ class Output:
 
     def unprotect(self):
         if self.exists and self.use_cache:
-            self.cache.unprotect(self.fs_path)
+            with TqdmCallback(
+                size=self.meta.nfiles or -1, desc=f"Unprotecting {self}"
+            ) as callback:
+                self.cache.unprotect(self.fs_path, callback=callback)
 
     def get_dir_cache(self, **kwargs) -> Optional["Tree"]:
         if not self.is_dir_checksum:
@@ -1151,7 +1154,7 @@ class Output:
         assert obj is None or isinstance(obj, Tree)
         if filter_info and filter_info != self.fs_path:
             assert obj
-            prefix = self.fs.path.parts(self.fs.path.relpath(filter_info, self.fs_path))
+            prefix = self.fs.parts(self.fs.relpath(filter_info, self.fs_path))
             return obj.filter(prefix)
         return obj
 
@@ -1294,10 +1297,7 @@ class Output:
     def unstage(self, path: str) -> Tuple["Meta", "Tree"]:
         from pygtrie import Trie
 
-        from dvc_objects.fs.path import Path
-
-        assert isinstance(self.fs.path, Path)
-        rel_key = tuple(self.fs.path.parts(self.fs.path.relpath(path, self.fs_path)))
+        rel_key = tuple(self.fs.parts(self.fs.relpath(path, self.fs_path)))
 
         if self.hash_info:
             tree = self.get_dir_cache()
@@ -1330,11 +1330,8 @@ class Output:
     ) -> Tuple["Meta", "Tree"]:
         from pygtrie import Trie
 
-        from dvc_objects.fs.path import Path
-
-        assert isinstance(self.fs.path, Path)
         append_only = True
-        rel_key = tuple(self.fs.path.parts(self.fs.path.relpath(path, self.fs_path)))
+        rel_key = tuple(self.fs.parts(self.fs.relpath(path, self.fs_path)))
 
         if self.hash_info:
             tree = self.get_dir_cache()
@@ -1435,7 +1432,7 @@ class Output:
 
         assert staging
         assert obj.hash_info
-        with Callback.as_tqdm_callback(
+        with TqdmCallback(
             desc=f"Adding {self} to cache",
             unit="file",
         ) as cb:
@@ -1523,7 +1520,7 @@ CLOUD_SCHEMA = vol.All({str: {**META_SCHEMA, **CHECKSUMS_SCHEMA}}, vol.Length(ma
 ARTIFACT_SCHEMA: Dict[Any, Any] = {
     **CHECKSUMS_SCHEMA,
     **META_SCHEMA,
-    vol.Required(Output.PARAM_PATH): str,
+    Output.PARAM_PATH: str,
     Output.PARAM_PERSIST: bool,
     Output.PARAM_CLOUD: CLOUD_SCHEMA,
     Output.PARAM_HASH: str,

dvc/parsing/__init__.py

@@ -143,11 +143,11 @@ class DataResolver:
         self.parsing_config = repo.config.get("parsing", {})
 
         if os.path.isabs(wdir):
-            wdir = fs.path.relpath(wdir)
+            wdir = fs.relpath(wdir)
             wdir = "" if wdir == os.curdir else wdir
 
         self.wdir = wdir
-        self.relpath = fs.path.normpath(fs.path.join(self.wdir, "dvc.yaml"))
+        self.relpath = fs.normpath(fs.join(self.wdir, "dvc.yaml"))
 
         vars_ = d.get(VARS_KWD, [])
         check_interpolations(vars_, VARS_KWD, self.relpath)
@@ -248,7 +248,7 @@ class EntryDefinition:
             wdir = to_str(context.resolve_str(wdir))
         except (ContextError, ParseError) as exc:
             format_and_raise(exc, f"'{self.where}.{name}.wdir'", self.relpath)
-        return self.resolver.fs.path.join(self.wdir, wdir)
+        return self.resolver.fs.join(self.wdir, wdir)
 
     def resolve(self, **kwargs):
         try:

dvc/parsing/context.py

@@ -384,7 +384,7 @@ class Context(CtxDict):
 
     def merge_from(self, fs, item: str, wdir: str, overwrite=False):
         path, _, keys_str = item.partition(":")
-        path = fs.path.normpath(fs.path.join(wdir, path))
+        path = fs.normpath(fs.join(wdir, path))
 
         select_keys = lfilter(bool, keys_str.split(",")) if keys_str else None
         if path in self.imports:
@@ -429,7 +429,7 @@ class Context(CtxDict):
         default: Optional[str] = None,
     ):
         if default:
-            to_import = fs.path.join(wdir, default)
+            to_import = fs.join(wdir, default)
             if fs.exists(to_import):
                 self.merge_from(fs, default, wdir)
             else:

dvc/render/__init__.py

@@ -1,7 +1,8 @@
-INDEX_FIELD = "step"
-REVISION_FIELD = "rev"
-FILENAME_FIELD = "filename"
-VERSION_FIELD = "dvc_data_version_info"
-REVISIONS_KEY = "revisions"
+INDEX = "step"
+REVISION = "rev"
+FILENAME = "filename"
+FIELD = "field"
+REVISIONS = "revisions"
+ANCHOR_DEFINITIONS = "anchor_definitions"
 TYPE_KEY = "type"
-SRC_FIELD = "src"
+SRC = "src"

dvc/render/convert.py

@@ -1,7 +1,6 @@
-from collections import defaultdict
 from typing import Dict, List, Union
 
-from dvc.render import REVISION_FIELD, REVISIONS_KEY, SRC_FIELD, TYPE_KEY, VERSION_FIELD
+from dvc.render import REVISION, REVISIONS, SRC, TYPE_KEY
 from dvc.render.converter.image import ImageConverter
 from dvc.render.converter.vega import VegaConverter
 
@@ -19,39 +18,31 @@ def _get_converter(
     raise ValueError(f"Invalid renderer class {renderer_class}")
 
 
-def _group_by_rev(datapoints):
-    grouped = defaultdict(list)
-    for datapoint in datapoints:
-        rev = datapoint.get(VERSION_FIELD, {}).get("revision")
-        grouped[rev].append(datapoint)
-    return dict(grouped)
-
-
 def to_json(renderer, split: bool = False) -> List[Dict]:
     if renderer.TYPE == "vega":
-        grouped = _group_by_rev(renderer.datapoints)
+        if not renderer.datapoints:
+            return []
+        revs = renderer.get_revs()
         if split:
-            content = renderer.get_filled_template(
-                skip_anchors=["data"], as_string=False
-            )
+            content, split_content = renderer.get_partial_filled_template()
         else:
-            content = renderer.get_filled_template(as_string=False)
-        if grouped:
-            return [
-                {
-                    TYPE_KEY: renderer.TYPE,
-                    REVISIONS_KEY: sorted(grouped.keys()),
-                    "content": content,
-                    "datapoints": grouped,
-                }
-            ]
-        return []
+            content = renderer.get_filled_template()
+            split_content = {}
+
+        return [
+            {
+                TYPE_KEY: renderer.TYPE,
+                REVISIONS: revs,
+                "content": content,
+                **split_content,
+            }
+        ]
     if renderer.TYPE == "image":
         return [
             {
                 TYPE_KEY: renderer.TYPE,
-                REVISIONS_KEY: [datapoint.get(REVISION_FIELD)],
-                "url": datapoint.get(SRC_FIELD),
+                REVISIONS: [datapoint.get(REVISION)],
+                "url": datapoint.get(SRC),
             }
             for datapoint in renderer.datapoints
         ]

dvc/render/converter/image.py

@@ -2,7 +2,7 @@ import base64
 import os
 from typing import TYPE_CHECKING, Any, Dict, List, Tuple
 
-from dvc.render import FILENAME_FIELD, REVISION_FIELD, SRC_FIELD
+from dvc.render import FILENAME, REVISION, SRC
 
 from . import Converter
 
@@ -58,9 +58,9 @@ class ImageConverter(Converter):
             else:
                 src = self._encode_image(image_data)
             datapoint = {
-                REVISION_FIELD: revision,
-                FILENAME_FIELD: filename,
-                SRC_FIELD: src,
+                REVISION: revision,
+                FILENAME: filename,
+                SRC: src,
             }
             datapoints.append(datapoint)
         return datapoints, properties

dvc/render/converter/vega.py

@@ -4,7 +4,7 @@ from typing import Any, Dict, Iterable, List, Optional, Tuple, Union
 from funcy import first, last
 
 from dvc.exceptions import DvcException
-from dvc.render import FILENAME_FIELD, INDEX_FIELD, VERSION_FIELD
+from dvc.render import FIELD, FILENAME, INDEX, REVISION
 
 from . import Converter
 
@@ -112,7 +112,6 @@ class VegaConverter(Converter):
     ):
         super().__init__(plot_id, data, properties)
         self.plot_id = plot_id
-        self.inferred_properties: Dict = {}
 
     def _infer_y_from_data(self):
         if self.plot_id in self.data:
@@ -120,34 +119,38 @@ class VegaConverter(Converter):
                 if all(isinstance(item, dict) for item in lst):
                     datapoint = first(lst)
                     field = last(datapoint.keys())
-                    self.inferred_properties["y"] = {self.plot_id: field}
-                    break
+                    return {self.plot_id: field}
+        return None
 
     def _infer_x_y(self):
         x = self.properties.get("x", None)
         y = self.properties.get("y", None)
 
+        inferred_properties: Dict = {}
+
         # Infer x.
         if isinstance(x, str):
-            self.inferred_properties["x"] = {}
+            inferred_properties["x"] = {}
             # If multiple y files, duplicate x for each file.
             if isinstance(y, dict):
                 for file, fields in y.items():
                     # Duplicate x for each y.
                     if isinstance(fields, list):
-                        self.inferred_properties["x"][file] = [x] * len(fields)
+                        inferred_properties["x"][file] = [x] * len(fields)
                     else:
-                        self.inferred_properties["x"][file] = x
+                        inferred_properties["x"][file] = x
             # Otherwise use plot ID as file.
             else:
-                self.inferred_properties["x"][self.plot_id] = x
+                inferred_properties["x"][self.plot_id] = x
 
         # Infer y.
         if y is None:
-            self._infer_y_from_data()
+            inferred_properties["y"] = self._infer_y_from_data()
         # If y files not provided, use plot ID as file.
         elif not isinstance(y, dict):
-            self.inferred_properties["y"] = {self.plot_id: y}
+            inferred_properties["y"] = {self.plot_id: y}
+
+        return inferred_properties
 
     def _find_datapoints(self):
         result = {}
@@ -182,7 +185,7 @@ class VegaConverter(Converter):
 
         x = properties.get("x", None)
         if not isinstance(x, dict):
-            return INDEX_FIELD
+            return INDEX
 
         fields = {field for _, field in _file_field(x)}
         if len(fields) == 1:
@@ -192,7 +195,7 @@ class VegaConverter(Converter):
     def flat_datapoints(self, revision):  # noqa: C901, PLR0912
         file2datapoints, properties = self.convert()
 
-        props_update = {}
+        props_update: Dict[str, Union[str, List[Dict[str, str]]]] = {}
 
         xs = list(_get_xs(properties, file2datapoints))
 
@@ -200,15 +203,17 @@ class VegaConverter(Converter):
         if not xs:
             x_file, x_field = (
                 None,
-                INDEX_FIELD,
+                INDEX,
             )
         else:
             x_file, x_field = xs[0]
-        props_update["x"] = x_field
+
+        num_xs = len(xs)
+        multiple_x_fields = num_xs > 1 and len({x[1] for x in xs}) > 1
+        props_update["x"] = "dvc_inferred_x_value" if multiple_x_fields else x_field
 
         ys = list(_get_ys(properties, file2datapoints))
 
-        num_xs = len(xs)
         num_ys = len(ys)
         if num_xs > 1 and num_xs != num_ys:
             raise DvcException(
@@ -237,6 +242,14 @@ class VegaConverter(Converter):
         else:
             common_prefix_len = 0
 
+        props_update["anchors_y_definitions"] = [
+            {
+                FILENAME: _get_short_y_file(y_file, common_prefix_len),
+                FIELD: y_field,
+            }
+            for y_file, y_field in ys
+        ]
+
         for i, (y_file, y_field) in enumerate(ys):
             if num_xs > 1:
                 x_file, x_field = xs[i]
@@ -249,15 +262,16 @@ class VegaConverter(Converter):
                     source_field=y_field,
                 )
 
-            if x_field == INDEX_FIELD and x_file is None:
-                _update_from_index(datapoints, INDEX_FIELD)
+            if x_field == INDEX and x_file is None:
+                _update_from_index(datapoints, INDEX)
             else:
                 x_datapoints = file2datapoints.get(x_file, [])
                 try:
                     _update_from_field(
                         datapoints,
-                        field=x_field,
+                        field="dvc_inferred_x_value" if multiple_x_fields else x_field,
                         source_datapoints=x_datapoints,
+                        source_field=x_field,
                     )
                 except IndexError:
                     raise DvcException(  # noqa: B904
@@ -266,15 +280,12 @@ class VegaConverter(Converter):
                         "They have to have same length."
                     )
 
-            y_file_short = y_file[common_prefix_len:].strip("/\\")
             _update_all(
                 datapoints,
                 update_dict={
-                    VERSION_FIELD: {
-                        "revision": revision,
-                        FILENAME_FIELD: y_file_short,
-                        "field": y_field,
-                    }
+                    REVISION: revision,
+                    FILENAME: _get_short_y_file(y_file, common_prefix_len),
+                    FIELD: y_field,
                 },
             )
 
@@ -295,10 +306,10 @@ class VegaConverter(Converter):
         generated datapoints and updated properties. `x`, `y` values and labels
         are inferred and always provided.
         """
-        self._infer_x_y()
+        inferred_properties = self._infer_x_y()
 
         datapoints = self._find_datapoints()
-        properties = {**self.properties, **self.inferred_properties}
+        properties = {**self.properties, **inferred_properties}
 
         properties["y_label"] = self.infer_y_label(properties)
         properties["x_label"] = self.infer_x_label(properties)
@@ -306,6 +317,10 @@ class VegaConverter(Converter):
         return datapoints, properties
 
 
+def _get_short_y_file(y_file, common_prefix_len):
+    return y_file[common_prefix_len:].strip("/\\")
+
+
 def _update_from_field(
     target_datapoints: List[Dict],
     field: str,

dvc/render/match.py

@@ -80,7 +80,7 @@ def match_defs_renderers(  # noqa: C901, PLR0912
     for plot_id, group in plots_data.group_definitions().items():
         plot_datapoints: List[Dict] = []
         props = _squash_plots_properties(group)
-        final_props: Dict = {}
+        first_props: Dict = {}
 
         def_errors: Dict[str, Exception] = {}
         src_errors: DefaultDict[str, Dict[str, Exception]] = defaultdict(dict)
@@ -90,6 +90,7 @@ def match_defs_renderers(  # noqa: C901, PLR0912
         if templates_dir is not None:
             props["template_dir"] = templates_dir
 
+        revs = []
         for rev, inner_id, plot_definition in group:
             plot_sources = infer_data_sources(inner_id, plot_definition)
             definitions_data = plots_data.get_definition_data(plot_sources, rev)
@@ -109,19 +110,24 @@ def match_defs_renderers(  # noqa: C901, PLR0912
 
             try:
                 dps, rev_props = converter.flat_datapoints(rev)
+                if dps and rev not in revs:
+                    revs.append(rev)
             except Exception as e:  # noqa: BLE001
                 logger.warning("In %r, %s", rev, str(e).lower())
                 def_errors[rev] = e
                 continue
 
-            if not final_props and rev_props:
-                final_props = rev_props
+            if not first_props and rev_props:
+                first_props = rev_props
             plot_datapoints.extend(dps)
 
-        if "title" not in final_props:
-            final_props["title"] = renderer_id
+        if "title" not in first_props:
+            first_props["title"] = renderer_id
+
+        if revs:
+            first_props["revs_with_datapoints"] = revs
 
         if renderer_cls is not None:
-            renderer = renderer_cls(plot_datapoints, renderer_id, **final_props)
+            renderer = renderer_cls(plot_datapoints, renderer_id, **first_props)
             renderers.append(RendererWithErrors(renderer, dict(src_errors), def_errors))
     return renderers

dvc/repo/__init__.py

@@ -70,12 +70,14 @@ class Repo:
     from dvc.repo.commit import commit  # type: ignore[misc]
     from dvc.repo.destroy import destroy  # type: ignore[misc]
     from dvc.repo.diff import diff  # type: ignore[misc]
+    from dvc.repo.du import du as _du  # type: ignore[misc]
     from dvc.repo.fetch import fetch  # type: ignore[misc]
     from dvc.repo.freeze import freeze, unfreeze  # type: ignore[misc]
     from dvc.repo.gc import gc  # type: ignore[misc]
     from dvc.repo.get import get as _get  # type: ignore[misc]
     from dvc.repo.get_url import get_url as _get_url  # type: ignore[misc]
     from dvc.repo.imp import imp  # type: ignore[misc]
+    from dvc.repo.imp_db import imp_db  # type: ignore[misc]
     from dvc.repo.imp_url import imp_url  # type: ignore[misc]
     from dvc.repo.install import install  # type: ignore[misc]
     from dvc.repo.ls import ls as _ls  # type: ignore[misc]
@@ -92,6 +94,7 @@ class Repo:
     from .cache import check_missing as cache_check_missing  # type: ignore[misc]
     from .data import status as data_status  # type: ignore[misc]
 
+    du = staticmethod(_du)
     ls = staticmethod(_ls)
     ls_url = staticmethod(_ls_url)
     get = staticmethod(_get)
@@ -111,7 +114,7 @@ class Repo:
         try:
             root_dir = self.find_root(root_dir, fs)
             fs = fs or localfs
-            dvc_dir = fs.path.join(root_dir, self.DVC_DIR)
+            dvc_dir = fs.join(root_dir, self.DVC_DIR)
         except NotDvcRepoError:
             if not uninitialized:
                 raise
@@ -204,7 +207,7 @@ class Repo:
                 self.fs.makedirs(self.tmp_dir, exist_ok=True)
 
                 self.lock = make_lock(
-                    self.fs.path.join(self.tmp_dir, "lock"),
+                    self.fs.join(self.tmp_dir, "lock"),
                     tmp_dir=self.tmp_dir,
                     hardlink_lock=self.config["core"].get("hardlink_lock", False),
                     friendly=True,
@@ -248,7 +251,7 @@ class Repo:
         )
 
     @cached_property
-    def local_dvc_dir(self):
+    def local_dvc_dir(self) -> Optional[str]:
         from dvc.fs import GitFileSystem, LocalFileSystem
 
         if not self.dvc_dir:
@@ -260,10 +263,10 @@ class Repo:
         if not isinstance(self.fs, GitFileSystem):
             return None
 
-        relparts = ()
+        relparts: Tuple[str, ...] = ()
         if self.root_dir != "/":
             # subrepo
-            relparts = self.fs.path.relparts(self.root_dir, "/")
+            relparts = self.fs.relparts(self.root_dir, "/")
 
         dvc_dir = os.path.join(
             self.scm.root_dir,
@@ -392,7 +395,7 @@ class Repo:
             index = subrepo.index.data[workspace]
         else:
             index = self.index.data[workspace]
-            key = self.fs.path.relparts(path, self.root_dir)
+            key = self.fs.relparts(path, self.root_dir)
 
         try:
             return index, index[key]
@@ -408,18 +411,18 @@ class Repo:
 
         fs = fs or localfs
         root = root or os.curdir
-        root_dir = fs.path.abspath(root)
+        root_dir = fs.abspath(root)
 
         if not fs.isdir(root_dir):
             raise NotDvcRepoError(f"directory '{root}' does not exist")
 
         while True:
-            dvc_dir = fs.path.join(root_dir, cls.DVC_DIR)
+            dvc_dir = fs.join(root_dir, cls.DVC_DIR)
             if fs.isdir(dvc_dir):
                 return root_dir
             if isinstance(fs, LocalFileSystem) and os.path.ismount(root_dir):
                 break
-            parent = fs.path.parent(root_dir)
+            parent = fs.parent(root_dir)
             if parent == root_dir:
                 break
             root_dir = parent
@@ -437,7 +440,7 @@ class Repo:
 
         fs = fs or localfs
         root_dir = cls.find_root(root, fs=fs)
-        return fs.path.join(root_dir, cls.DVC_DIR)
+        return fs.join(root_dir, cls.DVC_DIR)
 
     @staticmethod
     def init(root_dir=os.curdir, no_scm=False, force=False, subdir=False) -> "Repo":
@@ -446,7 +449,10 @@ class Repo:
         return init(root_dir=root_dir, no_scm=no_scm, force=force, subdir=subdir)
 
     def unprotect(self, target):
-        return self.cache.repo.unprotect(target)
+        from dvc.fs.callbacks import TqdmCallback
+
+        with TqdmCallback(desc=f"Unprotecting {target}") as callback:
+            return self.cache.repo.unprotect(target, callback=callback)
 
     def _ignore(self):
         flist = [self.config.files["local"]]
@@ -536,19 +542,19 @@ class Repo:
         # using `outs_graph` to ensure graph checks are run
         outs = outs or self.index.outs_graph
 
-        abs_path = self.fs.path.abspath(path)
+        abs_path = self.fs.abspath(path)
         fs_path = abs_path
 
         def func(out):
             def eq(one, two):
                 return one == two
 
-            match = eq if strict else out.fs.path.isin_or_eq
+            match = eq if strict else out.fs.isin_or_eq
 
             if out.protocol == "local" and match(fs_path, out.fs_path):
                 return True
 
-            if recursive and out.fs.path.isin(out.fs_path, fs_path):
+            if recursive and out.fs.isin(out.fs_path, fs_path):
                 return True
 
             return False
@@ -560,7 +566,7 @@ class Repo:
         return matched
 
     def is_dvc_internal(self, path):
-        path_parts = self.fs.path.normpath(path).split(self.fs.sep)
+        path_parts = self.fs.normpath(path).split(self.fs.sep)
         return self.DVC_DIR in path_parts
 
     @cached_property
@@ -607,10 +613,10 @@ class Repo:
         cache_dir = self.config["core"].get("site_cache_dir") or site_cache_dir()
 
         if isinstance(self.fs, GitFileSystem):
-            relparts = ()
+            relparts: Tuple[str, ...] = ()
             if self.root_dir != "/":
                 # subrepo
-                relparts = self.fs.path.relparts(self.root_dir, "/")
+                relparts = self.fs.relparts(self.root_dir, "/")
             root_dir = os.path.join(self.scm.root_dir, *relparts)
         else:
             root_dir = self.root_dir

dvc/repo/add.py

@@ -181,7 +181,7 @@ def _add_transfer(
 
 def _add(stage: "Stage", source: Optional[str] = None, no_commit: bool = False) -> None:
     out = stage.outs[0]
-    path = out.fs.path.abspath(source) if source else None
+    path = out.fs.abspath(source) if source else None
     try:
         stage.add_outs(path, no_commit=no_commit)
     except CacheLinkError:

dvc/repo/artifacts.py

@@ -210,7 +210,8 @@ class Artifacts:
     ) -> Tuple[int, str]:
         from dvc_studio_client.model_registry import get_download_uris
 
-        from dvc.fs import Callback, HTTPFileSystem, generic, localfs
+        from dvc.fs import HTTPFileSystem, generic, localfs
+        from dvc.fs.callbacks import TqdmCallback
 
         logger.debug("Trying to download artifact '%s' via studio", name)
         out = out or os.getcwd()
@@ -228,7 +229,7 @@ class Artifacts:
                 dvc_studio_config=dvc_studio_config,
                 **kwargs,
             ).items():
-                to_info = localfs.path.join(out, path)
+                to_info = localfs.join(out, path)
                 if localfs.exists(to_info) and not force:
                     hint = "\nTo override it, re-run with '--force'."
                     raise FileExistsLocallyError(  # noqa: TRY301
@@ -244,7 +245,7 @@ class Artifacts:
             ) from exc
         fs = HTTPFileSystem()
         jobs = jobs or fs.jobs
-        with Callback.as_tqdm_callback(
+        with TqdmCallback(
             desc=f"Downloading '{name}' from '{repo_url}'",
             unit="files",
         ) as cb:
@@ -253,7 +254,7 @@ class Artifacts:
                 fs, from_infos, localfs, to_infos, callback=cb, batch_size=jobs
             )
 
-        return len(to_infos), relpath(localfs.path.commonpath(to_infos))
+        return len(to_infos), relpath(localfs.commonpath(to_infos))
 
     @classmethod
     def get(  # noqa: PLR0913

dvc/repo/brancher.py

@@ -20,6 +20,7 @@ def brancher(
     all_tags=False,
     all_commits=False,
     all_experiments=False,
+    workspace=True,
     commit_date: Optional[str] = None,
     sha_only=False,
     num=1,
@@ -31,6 +32,7 @@ def brancher(
         all_branches (bool): iterate over all available branches.
         all_commits (bool): iterate over all commits.
         all_tags (bool): iterate over all available tags.
+        workspace (bool): include workspace.
         commit_date (str): Keep experiments from the commits after(include)
                             a certain date. Date must match the extended
                             ISO 8601 format (YYYY-MM-DD).
@@ -58,12 +60,12 @@ def brancher(
     from dvc.fs import LocalFileSystem
 
     repo_root_parts: Tuple[str, ...] = ()
-    if self.fs.path.isin(self.root_dir, self.scm.root_dir):
-        repo_root_parts = self.fs.path.relparts(self.root_dir, self.scm.root_dir)
+    if self.fs.isin(self.root_dir, self.scm.root_dir):
+        repo_root_parts = self.fs.relparts(self.root_dir, self.scm.root_dir)
 
     cwd_parts: Tuple[str, ...] = ()
-    if self.fs.path.isin(self.fs.path.getcwd(), self.scm.root_dir):
-        cwd_parts = self.fs.path.relparts(self.fs.path.getcwd(), self.scm.root_dir)
+    if self.fs.isin(self.fs.getcwd(), self.scm.root_dir):
+        cwd_parts = self.fs.relparts(self.fs.getcwd(), self.scm.root_dir)
 
     saved_fs = self.fs
     saved_root = self.root_dir
@@ -73,7 +75,8 @@ def brancher(
 
     logger.trace("switching fs to workspace")
     self.fs = LocalFileSystem(url=self.root_dir)
-    yield "workspace"
+    if workspace:
+        yield "workspace"
 
     revs = revs.copy() if revs else []
     if "workspace" in revs:
@@ -122,18 +125,18 @@ def _switch_fs(
     logger.trace("switching fs to revision %s", rev[:7])
     assert isinstance(repo.scm, Git)
     fs = GitFileSystem(scm=repo.scm, rev=rev)
-    root_dir = repo.fs.path.join("/", *repo_root_parts)
+    root_dir = repo.fs.join("/", *repo_root_parts)
     if not fs.exists(root_dir):
         raise NotDvcRepoError(f"Commit '{rev[:7]}' does not contain a DVC repo")
 
     repo.fs = fs
     repo.root_dir = root_dir
-    repo.dvc_dir = fs.path.join(root_dir, repo.DVC_DIR)
+    repo.dvc_dir = fs.join(root_dir, repo.DVC_DIR)
     repo._reset()
 
     if cwd_parts:
-        cwd = repo.fs.path.join("/", *cwd_parts)
-        repo.fs.path.chdir(cwd)
+        cwd = repo.fs.join("/", *cwd_parts)
+        repo.fs.chdir(cwd)
 
 
 @contextmanager
@@ -145,12 +148,12 @@ def switch(repo: "Repo", rev: str) -> Iterator[str]:
         rev = resolve_rev(repo.scm, rev)
 
     repo_root_parts: Tuple[str, ...] = ()
-    if repo.fs.path.isin(repo.root_dir, repo.scm.root_dir):
-        repo_root_parts = repo.fs.path.relparts(repo.root_dir, repo.scm.root_dir)
+    if repo.fs.isin(repo.root_dir, repo.scm.root_dir):
+        repo_root_parts = repo.fs.relparts(repo.root_dir, repo.scm.root_dir)
 
     cwd_parts: Tuple[str, ...] = ()
-    if repo.fs.path.isin(repo.fs.path.getcwd(), repo.scm.root_dir):
-        cwd_parts = repo.fs.path.relparts(repo.fs.path.getcwd(), repo.scm.root_dir)
+    if repo.fs.isin(repo.fs.getcwd(), repo.scm.root_dir):
+        cwd_parts = repo.fs.relparts(repo.fs.getcwd(), repo.scm.root_dir)
 
     saved_fs = repo.fs
     saved_root = repo.root_dir

dvc/repo/checkout.py

@@ -81,7 +81,7 @@ def _check_can_delete(
         if cache_fs.exists(cache_path):
             continue
 
-        entry_paths.append(fs.path.join(path, *(entry.key or ())))
+        entry_paths.append(fs.join(path, *(entry.key or ())))
 
     if not entry_paths:
         return
@@ -172,7 +172,7 @@ def checkout(  # noqa: C901
         )
 
         for out_path in out_paths:
-            if self.fs.path.isin_or_eq(dest_path, out_path):
+            if self.fs.isin_or_eq(dest_path, out_path):
                 failed.add(out_path)
 
     with ui.progress(
@@ -195,7 +195,7 @@ def checkout(  # noqa: C901
 
     typ_map = {ADD: "added", DELETE: "deleted", MODIFY: "modified"}
     for key, typ in out_changes.items():
-        out_path = self.fs.path.join(self.root_dir, *key)
+        out_path = self.fs.join(self.root_dir, *key)
 
         if out_path in failed:
             self.fs.remove(out_path, recursive=True)

dvc/repo/commit.py

@@ -1,4 +1,4 @@
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, Set
 
 from dvc import prompt
 
@@ -7,6 +7,7 @@ from .scm_context import scm_context
 
 if TYPE_CHECKING:
     from . import Repo
+    from .index import IndexView
 
 
 def _prepare_message(stage, changes):
@@ -88,14 +89,7 @@ def commit_2_to_3(repo: "Repo", dry: bool = False):
         outs_filter=lambda o: o.hash_name == "md5-dos2unix",
         recursive=True,
     )
-    migrated = set()
-    for out in view.outs:
-        dvcfile = out.stage.dvcfile.relpath
-        if isinstance(out.stage.dvcfile, ProjectFile):
-            lockfile = out.stage.dvcfile._lockfile.relpath
-            migrated.add(f"{dvcfile} ({lockfile})")
-        else:
-            migrated.add(dvcfile)
+    migrated = _migrateable_dvcfiles(view)
     if not migrated:
         ui.write("No DVC files in the repo to migrate to the 3.0 format.")
         return
@@ -110,11 +104,41 @@ def commit_2_to_3(repo: "Repo", dry: bool = False):
             for out in stage.filter_outs(filter_info)
             if outs_filter is not None and outs_filter(out)
         }
+        modified = False
         if outs:
             for out in outs:
                 out.update_legacy_hash_name(force=True)
+            modified = True
+        deps = {dep for dep in stage.deps if not stage.is_import and dep.is_in_repo}
+        if deps:
+            for dep in deps:
+                dep.update_legacy_hash_name(force=True)
+            modified = True
+        if modified:
             stage.save(allow_missing=True)
             stage.commit(allow_missing=True, relink=True)
             if not isinstance(stage.dvcfile, ProjectFile):
                 ui.write(f"Updating DVC file '{stage.dvcfile.relpath}'")
             stage.dump(update_pipeline=False)
+
+
+def _migrateable_dvcfiles(view: "IndexView") -> Set[str]:
+    from dvc.dvcfile import ProjectFile
+
+    migrated = set()
+    for stage, filter_info in view._stage_infos:
+        outs_filter = view._outs_filter
+        dvcfile = stage.dvcfile.relpath
+        assert outs_filter
+        if any(outs_filter(out) for out in stage.filter_outs(filter_info)) or (
+            not stage.is_import
+            and any(
+                dep.is_in_repo and dep.hash_name == "md5-dos2unix" for dep in stage.deps
+            )
+        ):
+            if isinstance(stage.dvcfile, ProjectFile):
+                lockfile = stage.dvcfile._lockfile.relpath
+                migrated.add(f"{dvcfile} ({lockfile})")
+            else:
+                migrated.add(dvcfile)
+    return migrated

dvc/repo/data.py

@@ -205,7 +205,7 @@ class Status(TypedDict):
 
 def _transform_git_paths_to_dvc(repo: "Repo", files: Iterable[str]) -> List[str]:
     """Transform files rel. to Git root to DVC root, and drop outside files."""
-    rel = repo.fs.path.relpath(repo.root_dir, repo.scm.root_dir).rstrip("/")
+    rel = repo.fs.relpath(repo.root_dir, repo.scm.root_dir).rstrip("/")
 
     # if we have repo root in a different location than scm's root,
     # i.e. subdir repo, all git_paths need to be transformed rel. to the DVC
@@ -215,11 +215,11 @@ def _transform_git_paths_to_dvc(repo: "Repo", files: Iterable[str]) -> List[str]
         length = len(prefix)
         files = (file[length:] for file in files if file.startswith(prefix))
 
-    start = repo.fs.path.relpath(repo.fs.path.getcwd(), repo.root_dir)
+    start = repo.fs.relpath(repo.fs.getcwd(), repo.root_dir)
     if start in (os.curdir, ""):
         return list(files)
     # we need to convert repo relative paths to curdir relative.
-    return [repo.fs.path.relpath(file, start) for file in files]
+    return [repo.fs.relpath(file, start) for file in files]
 
 
 def status(

dvc/repo/du.py

@@ -0,0 +1,42 @@
+from typing import Any, Dict, Optional, Union
+
+
+def du(
+    url: str,
+    path: Optional[str] = None,
+    rev: Optional[str] = None,
+    summarize: bool = False,
+    config: Union[None, Dict[str, Any], str] = None,
+    remote: Optional[str] = None,
+    remote_config: Optional[dict] = None,
+):
+    from dvc.config import Config
+
+    from . import Repo
+
+    if config and not isinstance(config, dict):
+        config_dict = Config.load_file(config)
+    else:
+        config_dict = None
+
+    with Repo.open(
+        url,
+        rev=rev,
+        subrepos=True,
+        uninitialized=True,
+        config=config_dict,
+        remote=remote,
+        remote_config=remote_config,
+    ) as repo:
+        path = path or ""
+
+        fs = repo.dvcfs
+
+        if summarize or not fs.isdir(path):
+            return [(path, fs.du(path, total=True))]
+
+        ret = [
+            (entry_path, fs.du(entry_path, total=True)) for entry_path in fs.ls(path)
+        ]
+        ret.append((path, sum(entry[1] for entry in ret)))
+        return ret

dvc/repo/experiments/pull.py

@@ -112,4 +112,6 @@ def _pull_cache(
         refs = [refs]
     revs = list(exp_commits(repo.scm, refs))
     logger.debug("dvc fetch experiment '%s'", refs)
-    repo.fetch(jobs=jobs, remote=dvc_remote, run_cache=run_cache, revs=revs)
+    repo.fetch(
+        jobs=jobs, remote=dvc_remote, run_cache=run_cache, revs=revs, workspace=False
+    )

dvc/repo/experiments/push.py

@@ -188,4 +188,6 @@ def _push_cache(
     assert isinstance(repo.scm, Git)
     revs = list(exp_commits(repo.scm, refs))
     logger.debug("dvc push experiment '%s'", refs)
-    return repo.push(jobs=jobs, remote=dvc_remote, run_cache=run_cache, revs=revs)
+    return repo.push(
+        jobs=jobs, remote=dvc_remote, run_cache=run_cache, revs=revs, workspace=False
+    )

dvc/repo/experiments/stash.py

@@ -171,7 +171,7 @@ class ApplyStash(Stash):
         right_fs = self.scm.get_fs(right_rev)
         paths = [path for path in left_fs.find("/") if not right_fs.exists(path)]
         dest_paths = [
-            localfs.path.join(self.scm.root_dir, left_fs.path.relpath(path, "/"))
+            localfs.join(self.scm.root_dir, left_fs.relpath(path, "/"))
             for path in paths
         ]
         for src, dest in zip(paths, dest_paths):

dvc/repo/fetch.py

@@ -1,4 +1,4 @@
-from typing import List, Tuple
+from typing import TYPE_CHECKING, List, Tuple
 
 from dvc.exceptions import DownloadError
 from dvc.log import logger
@@ -7,6 +7,10 @@ from dvc_data.index import DataIndex, FileStorage
 
 from . import locked
 
+if TYPE_CHECKING:
+    from dvc.output import Output
+    from dvc.stage import Stage
+
 logger = logger.getChild(__name__)
 
 
@@ -28,10 +32,12 @@ def _collect_indexes(  # noqa: PLR0913
     recursive=False,
     all_commits=False,
     revs=None,
+    workspace=True,
     max_size=None,
     types=None,
     config=None,
     onerror=None,
+    push=False,
 ):
     indexes = {}
     collection_exc = None
@@ -42,11 +48,22 @@ def _collect_indexes(  # noqa: PLR0913
         core["remote"] = remote
         config["core"] = core
 
+    def stage_filter(stage: "Stage") -> bool:
+        if push and stage.is_repo_import:
+            return False
+        return True
+
+    def outs_filter(out: "Output") -> bool:
+        if push and not out.can_push:
+            return False
+        return True
+
     for rev in repo.brancher(
         revs=revs,
         all_branches=all_branches,
         all_tags=all_tags,
         all_commits=all_commits,
+        workspace=workspace,
     ):
         try:
             repo.config.merge(config)
@@ -57,16 +74,18 @@ def _collect_indexes(  # noqa: PLR0913
                 recursive=recursive,
                 max_size=max_size,
                 types=types,
+                stage_filter=stage_filter,
+                outs_filter=outs_filter,
             )
 
             idx.data["repo"].onerror = _make_index_onerror(onerror, rev)
 
             indexes[rev or "workspace"] = idx
-        except Exception as exc:
+        except Exception as exc:  # noqa: BLE001
             if onerror:
                 onerror(rev, None, exc)
             collection_exc = exc
-            logger.exception("failed to collect '%s'", rev or "workspace")
+            logger.warning("failed to collect '%s', skipping", rev or "workspace")
 
     if not indexes and collection_exc:
         raise collection_exc
@@ -87,6 +106,7 @@ def fetch(  # noqa: PLR0913
     all_commits=False,
     run_cache=False,
     revs=None,
+    workspace=True,
     max_size=None,
     types=None,
     config=None,
@@ -131,6 +151,7 @@ def fetch(  # noqa: PLR0913
         recursive=recursive,
         all_commits=all_commits,
         revs=revs,
+        workspace=workspace,
         max_size=max_size,
         types=types,
         config=config,
@@ -198,7 +219,7 @@ def _log_unversioned(data: List["DataIndex"]) -> Tuple[List["DataIndex"], int]:
         index.storage_map = fs_index.storage_map
         for key, entry in fs_index.iteritems():
             if entry.meta and not entry.meta.isdir and entry.meta.version_id is None:
-                unversioned.append(fs.unstrip_protocol(fs.path.join(remote.path, *key)))
+                unversioned.append(fs.unstrip_protocol(fs.join(remote.path, *key)))
             else:
                 index[key] = entry
         fs_index.close()

dvc/repo/graph.py

@@ -126,7 +126,7 @@ def build_graph(stages, outs_trie=None):
     outs_trie = outs_trie or build_outs_trie(stages)
 
     for stage in stages:
-        out = outs_trie.shortest_prefix(localfs.path.parts(stage.path)).value
+        out = outs_trie.shortest_prefix(localfs.parts(stage.path)).value
         if out:
             raise StagePathAsOutputError(stage, str(out))
 
@@ -135,9 +135,11 @@ def build_graph(stages, outs_trie=None):
     for stage in stages:
         if stage.is_repo_import:
             continue
+        if stage.is_db_import:
+            continue
 
         for dep in stage.deps:
-            dep_key = dep.fs.path.parts(dep.fs_path)
+            dep_key = dep.fs.parts(dep.fs_path)
             overlapping = [n.value for n in outs_trie.prefixes(dep_key)]
             if outs_trie.has_subtrie(dep_key):
                 overlapping.extend(outs_trie.values(prefix=dep_key))
@@ -160,8 +162,10 @@ def build_outs_graph(graph, outs_trie):
     for stage in graph.nodes():
         if stage.is_repo_import:
             continue
+        if stage.is_db_import:
+            continue
         for dep in stage.deps:
-            dep_key = dep.fs.path.parts(dep.fs_path)
+            dep_key = dep.fs.parts(dep.fs_path)
             overlapping = [n.value for n in outs_trie.prefixes(dep_key)]
             if outs_trie.has_subtrie(dep_key):
                 overlapping.extend(outs_trie.values(prefix=dep_key))

dvc/repo/imp_db.py

@@ -0,0 +1,81 @@
+from typing import TYPE_CHECKING, Any, Dict, Optional
+
+from dvc.exceptions import OutputDuplicationError
+from dvc.repo.scm_context import scm_context
+from dvc.ui import ui
+from dvc.utils import resolve_output, resolve_paths
+
+if TYPE_CHECKING:
+    from . import Repo
+
+from . import locked
+
+
+@locked
+@scm_context
+def imp_db(  # noqa: PLR0913
+    self: "Repo",
+    url: Optional[str] = None,
+    rev: Optional[str] = None,
+    project_dir: Optional[str] = None,
+    sql: Optional[str] = None,
+    model: Optional[str] = None,
+    version: Optional[int] = None,
+    frozen: bool = True,
+    profile: Optional[str] = None,
+    target: Optional[str] = None,
+    output_format: str = "csv",
+    out: Optional[str] = None,
+    force: bool = False,
+    connection: Optional[str] = None,
+):
+    ui.warn("WARNING: import-db is an experimental feature.")
+    ui.warn(
+        "The functionality may change or break without notice, "
+        "which could lead to unexpected behavior."
+    )
+    erepo = None
+    if model and url:
+        erepo = {"url": url}
+        if rev:
+            erepo["rev"] = rev
+
+    assert output_format in ("csv", "json")
+
+    db: Dict[str, Any] = {"file_format": output_format}
+    if profile:
+        db["profile"] = profile
+
+    if model:
+        out = out or f"{model}.{output_format}"
+        db.update({"model": model, "version": version, "project_dir": project_dir})
+    else:
+        out = out or f"results.{output_format}"
+        db.update({"query": sql, "connection": connection})
+
+    out = resolve_output(url or ".", out, force=force)
+    path, wdir, out = resolve_paths(self, out, always_local=True)
+    stage = self.stage.create(
+        single_stage=True,
+        validate=False,
+        fname=path,
+        wdir=wdir,
+        deps=[url],
+        outs=[out],
+        erepo=erepo,
+        fs_config=None,
+        db=db,
+    )
+
+    stage.deps[0].target = target
+    try:
+        self.check_graph(stages={stage})
+    except OutputDuplicationError as exc:
+        raise OutputDuplicationError(  # noqa: B904
+            exc.output, set(exc.stages) - {stage}
+        )
+
+    stage.run()
+    stage.frozen = frozen
+    stage.dump()
+    return stage

dvc/repo/imports.py

@@ -100,10 +100,10 @@ def save_imports(
     Returns:
         Objects which were downloaded from source location.
     """
+    from dvc.fs.callbacks import TqdmCallback
     from dvc.stage.exceptions import DataSourceChanged
     from dvc_data.index import md5, save
     from dvc_data.index.checkout import apply, compare
-    from dvc_objects.fs.callbacks import Callback
 
     downloaded: Set["HashInfo"] = set()
 
@@ -122,7 +122,7 @@ def save_imports(
             if not cache.fs.exists(cache.path):
                 os.makedirs(cache.path)
             with TemporaryDirectory(dir=cache.path) as tmpdir:
-                with Callback.as_tqdm_callback(
+                with TqdmCallback(
                     desc="Downloading imports from source",
                     unit="files",
                 ) as cb:

dvc/repo/index.py

@@ -2,6 +2,7 @@ import logging
 import time
 from collections import defaultdict
 from functools import partial
+from itertools import chain
 from typing import (
     TYPE_CHECKING,
     Any,
@@ -19,6 +20,7 @@ from typing import (
 
 from funcy.debug import format_time
 
+from dvc.dependency import ParamsDependency
 from dvc.fs import LocalFileSystem
 from dvc.fs.callbacks import DEFAULT_CALLBACK
 from dvc.log import logger
@@ -28,7 +30,7 @@ if TYPE_CHECKING:
     from networkx import DiGraph
     from pygtrie import Trie
 
-    from dvc.dependency import Dependency, ParamsDependency
+    from dvc.dependency import Dependency
     from dvc.fs.callbacks import Callback
     from dvc.output import Output
     from dvc.repo import Repo
@@ -95,7 +97,7 @@ def collect_files(
     for root, dirs, files in walk_iter:
         dvcfile_filter = partial(is_dvcfile_and_not_ignored, root)
         for file in filter(dvcfile_filter, files):
-            file_path = fs.path.join(root, file)
+            file_path = fs.join(root, file)
             try:
                 index = Index.from_file(repo, file_path)
             except DvcException as exc:
@@ -133,7 +135,11 @@ def _load_data_from_outs(index, prefix, outs):
             hash_info=out.hash_info,
         )
 
-        if out.stage.is_import and not out.stage.is_repo_import:
+        if (
+            out.stage.is_import
+            and not out.stage.is_repo_import
+            and not out.stage.is_db_import
+        ):
             dep = out.stage.deps[0]
             entry.meta = dep.meta
             if out.hash_info:
@@ -181,6 +187,9 @@ def _load_storage_from_out(storage_map, key, out):
     except NoRemoteError:
         pass
 
+    if out.stage.is_db_import:
+        return
+
     if out.stage.is_import:
         dep = out.stage.deps[0]
         if not out.hash_info:
@@ -192,7 +201,7 @@ def _load_storage_from_out(storage_map, key, out):
                 FileStorage(
                     key,
                     fs_cache.fs,
-                    fs_cache.fs.path.join(
+                    fs_cache.fs.join(
                         fs_cache.path, dep.fs.protocol, tokenize(dep.fs_path)
                     ),
                 )
@@ -405,7 +414,25 @@ class Index:
             by_workspace[workspace].add(key)
 
         for path in _collect_top_level_metrics(self.repo):
-            key = self.repo.fs.path.relparts(path, self.repo.root_dir)
+            key = self.repo.fs.relparts(path, self.repo.root_dir)
+            by_workspace["repo"].add(key)
+
+        return dict(by_workspace)
+
+    @cached_property
+    def param_keys(self) -> Dict[str, Set["DataIndexKey"]]:
+        from .params.show import _collect_top_level_params
+
+        by_workspace: Dict[str, Set["DataIndexKey"]] = defaultdict(set)
+        by_workspace["repo"] = set()
+
+        param_paths = _collect_top_level_params(self.repo)
+        default_file: str = ParamsDependency.DEFAULT_PARAMS_FILE
+        if self.repo.fs.exists(f"{self.repo.fs.root_marker}{default_file}"):
+            param_paths = chain(param_paths, [default_file])
+
+        for path in param_paths:
+            key = self.repo.fs.relparts(path, self.repo.root_dir)
             by_workspace["repo"].add(key)
 
         return dict(by_workspace)
@@ -424,7 +451,7 @@ class Index:
             by_workspace[workspace].add(key)
 
         for path in self._plot_sources:
-            key = self.repo.fs.path.parts(path)
+            key = self.repo.fs.parts(path)
             by_workspace["repo"].add(key)
 
         return dict(by_workspace)
@@ -550,6 +577,8 @@ class Index:
                 keys = self.plot_keys
             elif typ == "metrics":
                 keys = self.metric_keys
+            elif typ == "params":
+                keys = self.param_keys
             else:
                 raise ValueError(f"unsupported type {typ}")
 
@@ -672,8 +701,8 @@ class IndexView:
             if not out.use_cache:
                 continue
             workspace, key = out.index_key
-            if filter_info and out.fs.path.isin(filter_info, out.fs_path):
-                key = key + out.fs.path.relparts(filter_info, out.fs_path)
+            if filter_info and out.fs.isin(filter_info, out.fs_path):
+                key = key + out.fs.relparts(filter_info, out.fs_path)
             entry = self._index.data[workspace].get(key)
             if entry and entry.meta and entry.meta.isdir:
                 prefixes[workspace].recursive.add(key)
@@ -689,8 +718,8 @@ class IndexView:
                 continue
 
             workspace, key = out.index_key
-            if filter_info and out.fs.path.isin(filter_info, out.fs_path):
-                key = key + out.fs.path.relparts(filter_info, out.fs_path)
+            if filter_info and out.fs.isin(filter_info, out.fs_path):
+                key = key + out.fs.relparts(filter_info, out.fs_path)
             ret[workspace].add(key)
 
         return dict(ret)
@@ -753,7 +782,7 @@ def build_data_index(  # noqa: C901, PLR0912
     data = DataIndex()
     parents = set()
     for key in index.data_keys.get(workspace, set()):
-        out_path = fs.path.join(path, *key)
+        out_path = fs.join(path, *key)
 
         for key_len in range(1, len(key)):
             parents.add(key[:key_len])
@@ -798,7 +827,7 @@ def build_data_index(  # noqa: C901, PLR0912
             callback.relative_update(1)
 
     for key in parents:
-        parent_path = fs.path.join(path, *key)
+        parent_path = fs.join(path, *key)
         if not fs.exists(parent_path):
             continue
         direntry = DataIndexEntry(key=key, meta=Meta(isdir=True), loaded=True)

dvc/repo/ls.py

@@ -89,7 +89,7 @@ def _ls(
         if not recursive:
             files.update(dirs)
 
-        parts = fs.path.relparts(root, fs_path)
+        parts = fs.relparts(root, fs_path)
         if parts == (".",):
             parts = ()
 

dvc/repo/ls_url.py

@@ -18,7 +18,7 @@ def ls_url(url, *, fs_config=None, recursive=False, config=None):
 
         for info in files.values():
             ls_info = {
-                "path": fs.path.relpath(info["name"], fs_path),
+                "path": fs.relpath(info["name"], fs_path),
                 "isdir": info["type"] == "directory",
                 "size": info.get("size"),
             }

dvc/repo/metrics/show.py

@@ -35,10 +35,10 @@ logger = logger.getChild(__name__)
 def _collect_top_level_metrics(repo: "Repo") -> Iterator[str]:
     top_metrics = repo.index._metrics
     for dvcfile, metrics in top_metrics.items():
-        wdir = repo.fs.path.relpath(repo.fs.path.parent(dvcfile), repo.root_dir)
+        wdir = repo.fs.relpath(repo.fs.parent(dvcfile), repo.root_dir)
         for file in metrics:
-            path = repo.fs.path.join(wdir, as_posix(file))
-            yield repo.fs.path.normpath(path)
+            path = repo.fs.join(wdir, as_posix(file))
+            yield repo.fs.normpath(path)
 
 
 def _extract_metrics(metrics, path: str):
@@ -101,7 +101,7 @@ def _collect_metrics(
 
     if not targets or outs_only:
         outs = metrics_from_target(repo, stages) if stages else repo.index.metrics
-        relpath = repo.fs.path.relpath
+        relpath = repo.fs.relpath
         metrics.extend(relpath(out.fs_path, repo.root_dir) for out in outs)
 
     if not targets and not outs_only and not stages:
@@ -128,8 +128,8 @@ class Result(TypedDict, total=False):
 
 
 def to_relpath(fs: "FileSystem", root_dir: str, d: Result) -> Result:
-    relpath = fs.path.relpath
-    cwd = fs.path.getcwd()
+    relpath = fs.relpath
+    cwd = fs.getcwd()
 
     start = relpath(cwd, root_dir)
     data = d.get("data")
@@ -156,7 +156,7 @@ def _gather_metrics(
     fs = repo.dvcfs
     for fs_path, result in _read_metrics(fs, files, cache=True):
         repo_path = fs_path.lstrip(fs.root_marker)
-        repo_os_path = os.sep.join(fs.path.parts(repo_path))
+        repo_os_path = os.sep.join(fs.parts(repo_path))
         if not isinstance(result, Exception):
             data.update({repo_os_path: FileResult(data=result)})
             continue

dvc/repo/open_repo.py

@@ -71,7 +71,7 @@ def erepo_factory(url, root_dir, cache_config):
         _config = cache_config.copy()
         if os.path.isdir(url):
             fs = fs or localfs
-            repo_path = os.path.join(url, *fs.path.relparts(path, root_dir))
+            repo_path = os.path.join(url, *fs.relparts(path, root_dir))
             _config.update(_get_remote_config(repo_path))
         return Repo(path, fs=fs, config=_config, **_kwargs)
 

dvc/repo/params/show.py

@@ -21,10 +21,10 @@ logger = logger.getChild(__name__)
 def _collect_top_level_params(repo: "Repo") -> Iterator[str]:
     top_params = repo.index._params
     for dvcfile, params in top_params.items():
-        wdir = repo.fs.path.relpath(repo.fs.path.parent(dvcfile), repo.root_dir)
+        wdir = repo.fs.relpath(repo.fs.parent(dvcfile), repo.root_dir)
         for file in params:
-            path = repo.fs.path.join(wdir, as_posix(file))
-            yield repo.fs.path.normpath(path)
+            path = repo.fs.join(wdir, as_posix(file))
+            yield repo.fs.normpath(path)
 
 
 def params_from_target(
@@ -55,7 +55,7 @@ def _collect_params(
 
     if not targets or stages:
         deps = params_from_target(repo, stages) if stages else repo.index.params
-        relpath = repo.fs.path.relpath
+        relpath = repo.fs.relpath
         params.extend(
             {relpath(dep.fs_path, repo.root_dir): list(dep.params)} for dep in deps
         )
@@ -93,7 +93,7 @@ def _collect_vars(repo, params, stages=None) -> Dict:
                 # to reduce noise and duplication, they are skipped
 
                 # `file` is relative
-                abspath = repo.fs.path.abspath(file)
+                abspath = repo.fs.abspath(file)
                 repo_path = repo.dvcfs.from_os_path(abspath)
                 if repo_path in params:
                     continue
@@ -138,7 +138,7 @@ def _gather_params(
     fs = repo.dvcfs
     for fs_path, result in _read_params(fs, files_keypaths, cache=True):
         repo_path = fs_path.lstrip(fs.root_marker)
-        repo_os_path = os.sep.join(fs.path.parts(repo_path))
+        repo_os_path = os.sep.join(fs.parts(repo_path))
         if not isinstance(result, Exception):
             data.update({repo_os_path: FileResult(data=result)})
             continue

dvc/repo/plots/__init__.py

@@ -386,7 +386,7 @@ def _matches(targets, config_file, plot_id):
 
 
 def _normpath(path):
-    # TODO dvcfs.path.normopath normalizes to windows path on Windows
+    # TODO dvcfs.normopath normalizes to windows path on Windows
     # even though other methods work as expected
     import posixpath
 
@@ -399,7 +399,7 @@ def _relpath(fs, path):
     # and invoking from some subdir `dvcfile.relpath` returns strange long
     # relative paths
     # ("../../../../../../dvc.yaml") - investigate
-    return fs.path.relpath(fs.path.join("/", fs.from_os_path(path)), fs.path.getcwd())
+    return fs.relpath(fs.join("/", fs.from_os_path(path)), fs.getcwd())
 
 
 def _collect_output_plots(repo, targets, props, onerror: Optional[Callable] = None):
@@ -413,7 +413,7 @@ def _collect_output_plots(repo, targets, props, onerror: Optional[Callable] = No
         if _matches(targets, config_path, str(plot)):
             unpacked = unpack_if_dir(
                 fs,
-                _normpath(fs.path.join(wdir_relpath, plot.def_path)),
+                _normpath(fs.join(wdir_relpath, plot.def_path)),
                 props={**plot_props, **props},
                 onerror=onerror,
             )
@@ -438,7 +438,7 @@ def _adjust_sources(fs, plot_props, config_dir):
     old_y = new_plot_props.pop("y", {})
     new_y = {}
     for filepath, val in old_y.items():
-        new_y[_normpath(fs.path.join(config_dir, filepath))] = val
+        new_y[_normpath(fs.join(config_dir, filepath))] = val
     new_plot_props["y"] = new_y
     return new_plot_props
 
@@ -452,13 +452,13 @@ def _resolve_definitions(
     onerror: Optional[Callable[[Any], Any]] = None,
 ):
     config_path = os.fspath(config_path)
-    config_dir = fs.path.dirname(config_path)
+    config_dir = fs.dirname(config_path)
     result: Dict[str, Dict] = {}
     for plot_id, plot_props in definitions.items():
         if plot_props is None:
             plot_props = {}
         if _id_is_path(plot_props):
-            data_path = _normpath(fs.path.join(config_dir, plot_id))
+            data_path = _normpath(fs.join(config_dir, plot_id))
             if _matches(targets, config_path, plot_id):
                 unpacked = unpack_if_dir(
                     fs,

dvc/repo/push.py

@@ -46,6 +46,7 @@ def push(  # noqa: PLR0913
     all_commits=False,
     run_cache=False,
     revs=None,
+    workspace=True,
     glob=False,
 ):
     from fsspec.utils import tokenize
@@ -87,6 +88,8 @@ def push(  # noqa: PLR0913
         recursive=recursive,
         all_commits=all_commits,
         revs=revs,
+        workspace=workspace,
+        push=True,
     )
 
     cache_key = (
@@ -107,6 +110,7 @@ def push(  # noqa: PLR0913
             push=True,
         )
 
+    push_transferred, push_failed = 0, 0
     try:
         with ui.progress(
             desc="Pushing",
@@ -131,6 +135,11 @@ def push(  # noqa: PLR0913
         for fs_index in data:
             fs_index.close()
 
+        if push_transferred:
+            # NOTE: dropping cached index to force reloading from newly saved
+            # metadata from version-aware remotes
+            self.drop_data_index()
+
     transferred_count += push_transferred
     failed_count += push_failed
     if failed_count:

dvc/repo/stage.py

@@ -219,7 +219,7 @@ class StageLoad:
         self, path: Optional[str] = None, name: Optional[str] = None
     ) -> str:
         if path:
-            return self.repo.fs.path.abspath(path)
+            return self.repo.fs.abspath(path)
 
         path = PROJECT_FILE
         logger.debug("Assuming '%s' to be a stage inside '%s'", name, path)
@@ -348,7 +348,7 @@ class StageLoad:
         if recursive and self.fs.isdir(target):
             from dvc.repo.graph import collect_inside_path
 
-            path = self.fs.path.abspath(target)
+            path = self.fs.abspath(target)
             return collect_inside_path(path, graph or self.repo.index.graph)
 
         stages = self.from_target(target, glob=glob)
@@ -393,7 +393,7 @@ class StageLoad:
             if not (recursive and self.fs.isdir(target)):
                 try:
                     (out,) = self.repo.find_outs_by_path(target, strict=False)
-                    return [StageInfo(out.stage, self.fs.path.abspath(target))]
+                    return [StageInfo(out.stage, self.fs.abspath(target))]
                 except OutputNotFoundError:
                     pass
 

dvc/repo/trie.py

@@ -9,7 +9,7 @@ def build_outs_trie(stages):
 
     for stage in stages:
         for out in stage.outs:
-            out_key = out.fs.path.parts(out.fs_path)
+            out_key = out.fs.parts(out.fs_path)
 
             # Check for dup outs
             if out_key in outs:

dvc/repo/worktree.py

@@ -3,7 +3,6 @@ from typing import TYPE_CHECKING, Any, Dict, Iterable, Optional, Tuple, Union
 
 from funcy import first
 
-from dvc.fs.callbacks import Callback
 from dvc.log import logger
 from dvc.stage.exceptions import StageUpdateError
 
@@ -101,7 +100,7 @@ def _get_remote(
     return repo.cloud.get_remote(name, command)
 
 
-def _merge_push_meta(
+def _merge_push_meta(  # noqa: C901
     out: "Output",
     index: Union["DataIndex", "DataIndexView"],
     remote: Optional[str] = None,
@@ -116,7 +115,11 @@ def _merge_push_meta(
     from dvc_data.index.save import build_tree
 
     _, key = out.index_key
-    entry = index[key]
+
+    entry = index.get(key)
+    if entry is None:
+        return
+
     repo = out.stage.repo
     if out.isdir():
         old_tree = out.get_obj()
@@ -128,10 +131,11 @@ def _merge_push_meta(
             entries.append(entry)
             if entry.meta is not None and entry.meta.isdir:
                 continue
-            fs_path = repo.fs.path.join(repo.root_dir, *subkey)
-            meta, hash_info = old_tree.get(
-                repo.fs.path.relparts(fs_path, out.fs_path)
-            ) or (None, None)
+            fs_path = repo.fs.join(repo.root_dir, *subkey)
+            meta, hash_info = old_tree.get(repo.fs.relparts(fs_path, out.fs_path)) or (
+                None,
+                None,
+            )
             entry.hash_info = hash_info
             if entry.meta:
                 entry.meta.remote = remote
@@ -243,11 +247,12 @@ def _fetch_out_changes(
     remote_index: Union["DataIndex", "DataIndexView"],
     remote: "Remote",
 ):
+    from dvc.fs.callbacks import TqdmCallback
     from dvc_data.index.checkout import apply, compare
 
     old, new = _get_diff_indexes(out, local_index, remote_index)
 
-    with Callback.as_tqdm_callback(
+    with TqdmCallback(
         unit="entry",
         desc="Comparing indexes",
     ) as cb:
@@ -261,9 +266,7 @@ def _fetch_out_changes(
         )
 
     total = len(new)
-    with Callback.as_tqdm_callback(
-        unit="file", desc=f"Updating '{out}'", disable=total == 0
-    ) as cb:
+    with TqdmCallback(unit="file", desc=f"Updating '{out}'", disable=total == 0) as cb:
         cb.set_size(total)
         apply(
             diff,
@@ -352,11 +355,11 @@ def _get_update_diff_index(
             # downloading
             if out.isdir():
                 if not entry.meta.isdir:
-                    fs_path = repo.fs.path.join(repo.root_dir, *entry.key)
+                    fs_path = repo.fs.join(repo.root_dir, *entry.key)
                     tree = out.obj
                     assert isinstance(tree, Tree)
                     _, entry.hash_info = tree.get(  # type: ignore[misc]
-                        repo.fs.path.relparts(fs_path, out.fs_path)
+                        repo.fs.relparts(fs_path, out.fs_path)
                     )
             else:
                 entry.hash_info = out.hash_info

dvc/rwlock.py

@@ -43,10 +43,10 @@ class RWLockFileFormatError(DvcException):
 
 @contextmanager
 def _edit_rwlock(lock_dir, fs, hardlink):
-    path = fs.path.join(lock_dir, RWLOCK_FILE)
+    path = fs.join(lock_dir, RWLOCK_FILE)
 
     rwlock_guard = make_lock(
-        fs.path.join(lock_dir, RWLOCK_LOCK),
+        fs.join(lock_dir, RWLOCK_LOCK),
         tmp_dir=lock_dir,
         hardlink_lock=hardlink,
     )
@@ -82,7 +82,7 @@ def _check_blockers(tmp_dir, lock, info, *, mode, waiters):  # noqa: C901, PLR09
     to_release = defaultdict(list)
     for path, infos in lock[mode].items():
         for waiter_path in waiters:
-            if localfs.path.overlaps(waiter_path, path):
+            if localfs.overlaps(waiter_path, path):
                 break
         else:
             continue

dvc/scm.py

@@ -24,6 +24,8 @@ from dvc.progress import Tqdm
 if TYPE_CHECKING:
     from scmrepo.progress import GitProgressEvent
 
+    from dvc.fs import FileSystem
+
 
 class SCMError(DvcException):
     """Base class for source control management errors."""
@@ -262,3 +264,30 @@ def iter_revs(
 
     rev_resolver = partial(resolve_rev, scm)
     return group_by(rev_resolver, results)
+
+
+def lfs_prefetch(fs: "FileSystem", paths: List[str]):
+    from scmrepo.git.lfs import fetch as _lfs_fetch
+
+    from dvc.fs.dvc import DVCFileSystem
+    from dvc.fs.git import GitFileSystem
+
+    if isinstance(fs, DVCFileSystem) and isinstance(fs.repo.fs, GitFileSystem):
+        git_fs = fs.repo.fs
+        scm = fs.repo.scm
+        assert isinstance(scm, Git)
+    else:
+        return
+
+    try:
+        if "filter=lfs" not in git_fs.open(".gitattributes").read():
+            return
+    except OSError:
+        return
+    with TqdmGit(desc="Checking for Git-LFS objects") as pbar:
+        _lfs_fetch(
+            scm,
+            [git_fs.rev],
+            include=[(path if path.startswith("/") else f"/{path}") for path in paths],
+            progress=pbar.update_git,
+        )

dvc/stage/__init__.py

@@ -105,7 +105,7 @@ def create_stage(cls: Type[_T], repo, path, **kwargs) -> _T:
     fill_stage_outputs(stage, **kwargs)
     check_no_externals(stage)
     fill_stage_dependencies(
-        stage, **project(kwargs, ["deps", "erepo", "params", "fs_config"])
+        stage, **project(kwargs, ["deps", "erepo", "params", "fs_config", "db"])
     )
     check_circular_dependency(stage)
     check_duplicated_arguments(stage)
@@ -281,9 +281,24 @@ class Stage(params.StageParams):
 
         return isinstance(self.deps[0], RepoDependency)
 
+    @property
+    def is_db_import(self) -> bool:
+        if not self.is_import:
+            return False
+
+        from dvc.dependency import DbDependency, DbtDependency
+
+        return isinstance(self.deps[0], (DbDependency, DbtDependency))
+
     @property
     def is_versioned_import(self) -> bool:
-        return self.is_import and self.deps[0].fs.version_aware
+        from dvc.dependency import DbDependency, DbtDependency
+
+        return (
+            self.is_import
+            and not isinstance(self.deps[0], (DbDependency, DbtDependency))
+            and self.deps[0].fs.version_aware
+        )
 
     def short_description(self) -> Optional["str"]:
         desc: Optional["str"] = None
@@ -446,6 +461,9 @@ class Stage(params.StageParams):
     ) -> None:
         if not (self.is_repo_import or self.is_import):
             raise StageUpdateError(self.relpath)
+
+        # always force update DbDep/DbtDep since we don't know if it's changed
+        force = self.is_db_import
         update_import(
             self,
             rev=rev,
@@ -453,6 +471,7 @@ class Stage(params.StageParams):
             remote=remote,
             no_download=no_download,
             jobs=jobs,
+            force=force,
         )
 
     def reload(self) -> "Stage":
@@ -637,7 +656,7 @@ class Stage(params.StageParams):
 
     def filter_outs(self, fs_path) -> Iterable["Output"]:
         def _func(o):
-            return o.fs.path.isin_or_eq(fs_path, o.fs_path)
+            return o.fs.isin_or_eq(fs_path, o.fs_path)
 
         return filter(_func, self.outs) if fs_path else self.outs
 

dvc/stage/cache.py

@@ -181,9 +181,9 @@ class StageCache:
 
         path = self._get_cache_path(cache_key, cache_value)
         local_fs = self.repo.cache.legacy.fs
-        parent = local_fs.path.parent(path)
+        parent = local_fs.parent(path)
         self.repo.cache.legacy.makedirs(parent)
-        tmp = local_fs.path.join(parent, fs.utils.tmp_fname())
+        tmp = local_fs.join(parent, fs.utils.tmp_fname())
         assert os.path.exists(parent)
         assert os.path.isdir(parent)
         dump_yaml(tmp, cache)
@@ -231,12 +231,12 @@ class StageCache:
 
     def transfer(self, from_odb, to_odb):
         from dvc.fs import HTTPFileSystem, LocalFileSystem
-        from dvc.fs.callbacks import Callback
+        from dvc.fs.callbacks import TqdmCallback
 
         from_fs = from_odb.fs
         to_fs = to_odb.fs
         func = fs.generic.log_exceptions(fs.generic.copy)
-        runs = from_fs.path.join(from_odb.path, "runs")
+        runs = from_fs.join(from_odb.path, "runs")
 
         http_odb = next(
             (odb for odb in (from_odb, to_odb) if isinstance(odb.fs, HTTPFileSystem)),
@@ -252,21 +252,21 @@ class StageCache:
             return ret
 
         for src in from_fs.find(runs):
-            rel = from_fs.path.relpath(src, from_odb.path)
+            rel = from_fs.relpath(src, from_odb.path)
             if not isinstance(to_fs, LocalFileSystem):
-                rel = from_fs.path.as_posix(rel)
+                rel = from_fs.as_posix(rel)
 
-            dst = to_fs.path.join(to_odb.path, rel)
-            key = to_fs.path.parent(dst)
+            dst = to_fs.join(to_odb.path, rel)
+            key = to_fs.parent(dst)
             # check if any build cache already exists for this key
             # TODO: check if MaxKeys=1 or something like that applies
             # or otherwise this will take a lot of time!
             if to_fs.exists(key) and first(to_fs.find(key)):
                 continue
 
-            src_name = from_fs.path.name(src)
-            parent_name = from_fs.path.name(from_fs.path.parent(src))
-            with Callback.as_tqdm_callback(
+            src_name = from_fs.name(src)
+            parent_name = from_fs.name(from_fs.parent(src))
+            with TqdmCallback(
                 desc=src_name,
                 bytes=True,
             ) as cb:

dvc/stage/decorators.py

@@ -7,6 +7,7 @@ from funcy import decorator
 def rwlocked(call, read=None, write=None):
     import sys
 
+    from dvc.dependency.db import AbstractDependency
     from dvc.dependency.repo import RepoDependency
     from dvc.rwlock import rwlock
 
@@ -27,7 +28,7 @@ def rwlocked(call, read=None, write=None):
             for item in getattr(stage, attr)
             # There is no need to lock RepoDependency deps, as there is no
             # corresponding OutputREPO, so we can't even write it.
-            if not isinstance(item, RepoDependency)
+            if not isinstance(item, (RepoDependency, AbstractDependency))
         ]
 
     cmd = " ".join(sys.argv)

dvc/stage/imports.py

@@ -18,7 +18,13 @@ def _update_import_on_remote(stage, remote, jobs):
 
 
 def update_import(
-    stage, rev=None, to_remote=False, remote=None, no_download=None, jobs=None
+    stage,
+    rev=None,
+    to_remote=False,
+    remote=None,
+    no_download=None,
+    jobs=None,
+    force=False,
 ):
     stage.deps[0].update(rev=rev)
 
@@ -30,7 +36,7 @@ def update_import(
         if to_remote:
             _update_import_on_remote(stage, remote, jobs)
         else:
-            stage.reproduce(no_download=no_download, jobs=jobs)
+            stage.reproduce(no_download=no_download, jobs=jobs, force=force)
     finally:
         if no_download and changed:
             # Avoid retaining stale information
@@ -55,7 +61,7 @@ def sync_import(
     else:
         stage.save_deps()
         if no_download:
-            if stage.is_repo_import:
+            if stage.is_repo_import or stage.is_db_import:
                 stage.deps[0].update()
         else:
             stage.deps[0].download(

dvc/stage/utils.py

@@ -77,12 +77,14 @@ def fill_stage_outputs(stage, **kwargs):
         )
 
 
-def fill_stage_dependencies(stage, deps=None, erepo=None, params=None, fs_config=None):
+def fill_stage_dependencies(
+    stage, deps=None, erepo=None, params=None, fs_config=None, db=None
+):
     from dvc.dependency import loads_from, loads_params
 
     assert not stage.deps
     stage.deps = []
-    stage.deps += loads_from(stage, deps or [], erepo=erepo, fs_config=fs_config)
+    stage.deps += loads_from(stage, deps or [], erepo=erepo, fs_config=fs_config, db=db)
     stage.deps += loads_params(stage, params or [])
 
 
@@ -182,9 +184,9 @@ def resolve_wdir(wdir, path):
 
 
 def resolve_paths(fs, path, wdir=None):
-    path = fs.path.abspath(path)
+    path = fs.abspath(path)
     wdir = wdir or os.curdir
-    wdir = fs.path.abspath(fs.path.join(fs.path.dirname(path), wdir))
+    wdir = fs.abspath(fs.join(fs.dirname(path), wdir))
     return path, wdir
 
 

dvc/testing/benchmarks/cli/commands/test_fetch.py

@@ -0,0 +1,3 @@
+def test_fetch(bench_dvc, tmp_dir, dvc, make_dataset, remote):
+    make_dataset(cache=False, dvcfile=True, files=False, remote=True)
+    bench_dvc("fetch")

dvc/testing/benchmarks/cli/commands/test_get.py

@@ -1,3 +1,7 @@
+import pytest
+
+
+@pytest.mark.flaky(max_runs=3, min_passes=1)
 def test_get(bench_dvc, tmp_dir, scm, dvc, make_dataset, remote):
     dataset = make_dataset(
         cache=False, files=False, dvcfile=True, commit=True, remote=True

dvc/testing/benchmarks/cli/commands/test_import.py

@@ -1,3 +1,7 @@
+import pytest
+
+
+@pytest.mark.flaky(max_runs=3, min_passes=1)
 def test_import(bench_dvc, tmp_dir, scm, dvc, make_dataset, remote):
     dataset = make_dataset(
         cache=False, files=False, dvcfile=True, commit=True, remote=True

dvc/testing/benchmarks/cli/commands/test_pull.py

@@ -1,3 +0,0 @@
-def test_pull(bench_dvc, tmp_dir, dvc, make_dataset, remote):
-    make_dataset(cache=False, dvcfile=True, files=False, remote=True)
-    bench_dvc("pull")

dvc/testing/benchmarks/cli/stories/use_cases/test_sharing.py

@@ -11,7 +11,8 @@ def test_sharing(bench_dvc, tmp_dir, dvc, dataset, remote):
     shutil.rmtree(dataset)
     shutil.rmtree(tmp_dir / ".dvc" / "cache")
 
-    bench_dvc("pull")
-    bench_dvc("pull", name="noop")
+    bench_dvc("fetch")
+    bench_dvc("fetch", name="noop")
 
+    bench_dvc("checkout")
     bench_dvc("checkout", name="noop")

dvc/testing/benchmarks/fixtures.py

@@ -2,26 +2,48 @@ import os
 import shutil
 from pathlib import Path
 from subprocess import check_output
+from typing import Dict, Optional
 
 import pytest
 from dulwich.porcelain import clone
 from funcy import first
 from packaging import version
 
+from dvc.types import StrPath
+
 
 @pytest.fixture(scope="session")
 def bench_config(request):
     return request.config.bench_config
 
 
+class VirtualEnv:
+    def __init__(self, path: StrPath) -> None:
+        self.path = Path(path)
+        self.bin = self.path / ("Scripts" if os.name == "nt" else "bin")
+
+    def create(self) -> None:
+        import virtualenv
+
+        virtualenv.cli_run([os.fspath(self.path)])
+
+    def run(self, cmd: str, *args: str, env: Optional[Dict[str, str]] = None) -> None:
+        exe = self.which(cmd)
+        check_output([exe, *args], env=env)  # noqa: S603
+
+    def which(self, cmd: str) -> str:
+        assert self.bin.exists()
+        return shutil.which(cmd, path=self.bin) or cmd
+
+
 @pytest.fixture(scope="session")
 def make_dvc_venv(tmp_path_factory):
     def _make_dvc_venv(name):
-        from pytest_virtualenv import VirtualEnv
-
         name = _sanitize_venv_name(name)
         venv_dir = tmp_path_factory.mktemp(f"dvc-venv-{name}")
-        return VirtualEnv(workspace=venv_dir)
+        venv = VirtualEnv(venv_dir)
+        venv.create()
+        return venv
 
     return _make_dvc_venv
 
@@ -74,16 +96,16 @@ def make_dvc_bin(
         venv = dvc_venvs.get(dvc_rev)
         if not venv:
             venv = make_dvc_venv(dvc_rev)
-            venv.run("pip install -U pip")
+            venv.run("pip", "install", "-U", "pip")
             if bench_config.dvc_install_deps:
                 egg = f"dvc[{bench_config.dvc_install_deps}]"
             else:
                 egg = "dvc"
-            venv.run(f"pip install git+file://{dvc_git_repo}@{dvc_rev}#egg={egg}")
+            venv.run("pip", "install", f"git+file://{dvc_git_repo}@{dvc_rev}#egg={egg}")
             if dvc_rev in ["2.18.1", "2.11.0", "2.6.3"]:
-                venv.run("pip install fsspec==2022.11.0")
+                venv.run("pip", "install", "fsspec==2022.11.0")
             dvc_venvs[dvc_rev] = venv
-        dvc_bin = venv.virtualenv / "bin" / "dvc"
+        dvc_bin = venv.which("dvc")
     else:
         dvc_bin = bench_config.dvc_bin
 
@@ -171,7 +193,7 @@ def make_dataset(request, bench_config, tmp_dir, dvc_bench_git_repo):
 
         path = tmp_dir / "dataset"
         root = dvc_bench_git_repo
-        src = root / "data" / bench_config.size / "dataset"
+        src = root / "data" / bench_config.dataset / "dataset"
         src_dvc = src.with_suffix(".dvc")
 
         dvc = Repo(root)

dvc/testing/benchmarks/plugin.py

@@ -4,6 +4,11 @@ DEFAULT_DVC_BENCH_GIT_REPO = "https://github.com/iterative/dvc-bench"
 DEFAULT_PROJECT_GIT_REPO = "https://github.com/iterative/example-get-started"
 
 
+def pytest_report_header(config):
+    bconf = config.bench_config
+    return "dvc-bench: (" f"dataset: '{bconf.dataset}'," f"revs: '{bconf.dvc_revs}'" ")"
+
+
 def pytest_generate_tests(metafunc):
     str_revs = metafunc.config.getoption("--dvc-revs")
     revs = str_revs.split(",") if str_revs else [None]
@@ -13,7 +18,7 @@ def pytest_generate_tests(metafunc):
 
 class DVCBenchConfig:
     def __init__(self):
-        self.size = "small"
+        self.dataset = "small"
         self.dvc_bin = DEFAULT_DVC_BIN
         self.dvc_revs = None
         self.dvc_git_repo = DEFAULT_DVC_GIT_REPO
@@ -29,7 +34,7 @@ def pytest_configure(config):
     )
 
     config.bench_config = DVCBenchConfig()
-    config.bench_config.size = config.getoption("--size")
+    config.bench_config.dataset = config.getoption("--dataset")
     config.bench_config.dvc_bin = config.getoption("--dvc-bin")
     config.bench_config.dvc_revs = config.getoption("--dvc-revs")
     config.bench_config.dvc_git_repo = config.getoption("--dvc-git-repo")
@@ -41,10 +46,10 @@ def pytest_configure(config):
 
 def pytest_addoption(parser):
     parser.addoption(
-        "--size",
-        choices=["tiny", "small", "large", "mnist"],
+        "--dataset",
+        type=str,
         default="small",
-        help="Size of the dataset/datafile to use in tests",
+        help="Dataset name to use in tests (e.g. tiny/small/large/mnist/etc)",
     )
 
     parser.addoption(

dvc/testing/remote_tests.py

@@ -172,6 +172,8 @@ class TestRemoteVersionAware:
         run_copy("foo", "foo_copy", name="copy")
 
         dvc.push()
+        assert (remote_version_aware / "foo").read_text() == "foo"
+        assert (remote_version_aware / "foo_copy").read_text() == "foo"
         foo_dvc = (tmp_dir / "foo.dvc").read_text()
         assert "version_id" in foo_dvc
         stage = stage.reload()
@@ -181,18 +183,24 @@ class TestRemoteVersionAware:
 
         remove(dvc.cache.local.path)
         remove(tmp_dir / "foo")
+        remove(tmp_dir / "foo_copy")
 
         dvc.pull()
         assert (tmp_dir / "foo").read_text() == "foo"
+        assert (tmp_dir / "foo_copy").read_text() == "foo"
         assert (tmp_dir / "foo.dvc").read_text() == foo_dvc
         assert (tmp_dir / "dvc.lock").read_text() == dvc_lock
 
         dvc.push()
+        assert (remote_version_aware / "foo").read_text() == "foo"
+        assert (remote_version_aware / "foo_copy").read_text() == "foo"
         assert (tmp_dir / "foo.dvc").read_text() == foo_dvc
         assert (tmp_dir / "dvc.lock").read_text() == dvc_lock
 
         dvc.reproduce()
         dvc.push()
+        assert (remote_version_aware / "foo").read_text() == "foo"
+        assert (remote_version_aware / "foo_copy").read_text() == "foo"
         assert (tmp_dir / "foo.dvc").read_text() == foo_dvc
         assert (tmp_dir / "dvc.lock").read_text() == dvc_lock
 
@@ -241,6 +249,12 @@ class TestRemoteVersionAware:
         assert (tmp_dir / "data_dir.dvc").read_text() == data_dir_dvc
         assert (tmp_dir / "dvc.lock").read_text() == dvc_lock
 
+        dvc.cache.local.clear()
+        remove(tmp_dir / "data_dir")
+        dvc.push()
+        assert (remote_version_aware / "data_dir").exists()
+        assert (remote_version_aware / "data_dir" / "data").exists()
+
 
 class TestRemoteWorktree:
     def test_file(self, tmp_dir, dvc, remote_worktree):

dvc/testing/tmp_dir.py

@@ -70,16 +70,20 @@ class TmpDir(pathlib.Path):
     def __new__(cls, *args, **kwargs):
         if cls is TmpDir:
             cls = WindowsTmpDir if os.name == "nt" else PosixTmpDir
-        # init parameter and `_init` method has been removed in Python 3.10.
-        kw = {"init": False} if sys.version_info < (3, 10) else {}
-        self = cls._from_parts(args, **kw)  # type: ignore[attr-defined]
-        if not self._flavour.is_supported:
-            raise NotImplementedError(
-                f"cannot instantiate {cls.__name__!r} on your system"
-            )
-        if sys.version_info < (3, 10):
-            self._init()
-        return self
+
+        if sys.version_info >= (3, 12):
+            return object.__new__(cls)
+        else:  # noqa: RET505
+            # init parameter and `_init` method has been removed in Python 3.10.
+            kw = {"init": False} if sys.version_info < (3, 10) else {}
+            self = cls._from_parts(args, **kw)  # type: ignore[attr-defined]
+            if not self._flavour.is_supported:
+                raise NotImplementedError(
+                    f"cannot instantiate {cls.__name__!r} on your system"
+                )
+            if sys.version_info < (3, 10):
+                self._init()
+            return self
 
     def init(self, *, scm=False, dvc=False, subdir=False):
         from dvc.repo import Repo

dvc/testing/workspace_tests.py

@@ -163,8 +163,8 @@ class TestImportURLVersionAware:
 
 
 def match_files(fs, entries, expected):
-    entries_content = {(fs.path.normpath(d["path"]), d["isdir"]) for d in entries}
-    expected_content = {(fs.path.normpath(d["path"]), d["isdir"]) for d in expected}
+    entries_content = {(fs.normpath(d["path"]), d["isdir"]) for d in entries}
+    expected_content = {(fs.normpath(d["path"]), d["isdir"]) for d in expected}
     assert entries_content == expected_content
 
 
@@ -177,7 +177,7 @@ class TestLsUrl:
         match_files(
             fs,
             result,
-            [{"path": fs.path.join(fs_path, fname), "isdir": False}],
+            [{"path": fs.join(fs_path, fname), "isdir": False}],
         )
 
     def test_dir(self, cloud):

dvc/ui/_rich_progress.py

@@ -4,6 +4,7 @@ from rich.progress import (
     DownloadColumn,
     MofNCompleteColumn,
     Progress,
+    SpinnerColumn,
     TextColumn,
     TimeElapsedColumn,
     TimeRemainingColumn,
@@ -52,3 +53,24 @@ class RichTransferProgress(RichProgress):
         yield self.make_tasks_table(summary_tasks)
         self.columns = self.TRANSFER_COLS
         yield self.make_tasks_table(other_tasks)
+
+
+class DbDownloadProgress(RichProgress):
+    PROGRESS_COLS = (
+        SpinnerColumn(),
+        TextColumn("[progress.description]{task.description}"),
+        TextColumn("[progress.download]{task.completed:,} rows"),
+        TextColumn("to [repr.filename]{task.fields[output]}"),
+    )
+    STATUS_COLS = (
+        SpinnerColumn(),
+        TextColumn("[progress.description]{task.description}"),
+        TextColumn("to [repr.filename]{task.fields[output]}"),
+    )
+
+    def get_renderables(self):
+        if self.tasks:
+            (task, *_) = self.tasks
+            cols = self.PROGRESS_COLS if task.completed else self.STATUS_COLS
+            self.columns = cols[1:] if task.finished else cols
+        yield self.make_tasks_table(self.tasks)

dvc/utils/__init__.py

@@ -287,7 +287,7 @@ def resolve_paths(repo, out, always_local=False):
         # urlparse interprets windows drive letters as URL scheme
         scheme = ""
 
-    if scheme or not localfs.path.isin_or_eq(abspath, repo.root_dir):
+    if scheme or not localfs.isin_or_eq(abspath, repo.root_dir):
         wdir = os.getcwd()
     elif contains_symlink_up_to(dirname, repo.root_dir) or (
         os.path.isdir(abspath) and localfs.is_symlink(abspath)

dvc/utils/packaging.py

@@ -0,0 +1,32 @@
+import logging
+
+from funcy import once_per_args
+
+from dvc.log import logger
+
+logger = logger.getChild(__name__)
+
+
+@once_per_args
+def check_required_version(pkg: str, dist: str = "dvc", log_level=logging.WARNING):
+    from importlib import metadata
+
+    from packaging.requirements import InvalidRequirement, Requirement
+
+    try:
+        reqs = {
+            r.name: r.specifier for r in map(Requirement, metadata.requires(dist) or [])
+        }
+        version = metadata.version(pkg)
+    except (metadata.PackageNotFoundError, InvalidRequirement):
+        return
+
+    specifier = reqs.get(pkg)
+    if specifier and version and version not in specifier:
+        logger.log(
+            log_level,
+            "%s%s is required, but you have %r installed which is incompatible.",
+            pkg,
+            specifier,
+            version,
+        )

dvc/utils/plots.py

@@ -9,7 +9,7 @@ def get_plot_id(config_plot_id: str, config_file_path: str = ""):
 
 
 def group_definitions_by_id(
-    definitions: Dict[str, Dict]
+    definitions: Dict[str, Dict],
 ) -> Dict[str, Tuple[str, Dict]]:
     """
     Format ID and extracts plot_definition for each plot.

dvc/utils/serialize/__init__.py

@@ -21,7 +21,7 @@ PARSERS.update(
 
 
 def load_path(fs_path, fs, **kwargs):
-    suffix = fs.path.suffix(fs_path).lower()
+    suffix = fs.suffix(fs_path).lower()
     loader = LOADERS[suffix]
     return loader(fs_path, fs=fs, **kwargs)
 

pyproject.toml

@@ -29,22 +29,23 @@ classifiers = [
     "Programming Language :: Python :: 3.11",
 ]
 dynamic = ["version"]
+
 dependencies = [
     "colorama>=0.3.9",
     "configobj>=5.0.6",
     "distro>=1.3",
     "dpath<3,>=2.1.0",
-    "dvc-data>=2.20.0,<2.21.0",
+    "dvc-data>=3,<3.1",
     "dvc-http>=2.29.0",
-    "dvc-render>=0.3.1,<1",
-    "dvc-studio-client>=0.13.0,<1",
+    "dvc-render>=1.0.0,<2",
+    "dvc-studio-client>=0.17.1,<1",
     "dvc-task>=0.3.0,<1",
     "flatten_dict<1,>=0.4.1",
     # https://github.com/iterative/dvc/issues/9654
     "flufl.lock>=5,<8",
     "funcy>=1.14",
     "grandalf<1,>=0.7",
-    "gto>=1.4.0,<2",
+    "gto>=1.6.0,<2",
     "hydra-core>=1.1",
     "iterative-telemetry>=0.0.7",
     "networkx>=2.5",
@@ -58,7 +59,7 @@ dependencies = [
     "requests>=2.22",
     "rich>=12",
     "ruamel.yaml>=0.17.11",
-    "scmrepo>=1.4.1,<2",
+    "scmrepo>=2.0.2,<3",
     "shortuuid>=0.5",
     "shtab<2,>=1.3.4",
     "tabulate>=0.8.7",
@@ -70,13 +71,14 @@ dependencies = [
 
 [project.optional-dependencies]
 all = ["dvc[azure,gdrive,gs,hdfs,oss,s3,ssh,webdav,webhdfs]"]
-azure = ["dvc-azure>=2.23.0"]
+azure = ["dvc-azure>=3,<4"]
 dev = ["dvc[azure,gdrive,gs,hdfs,lint,oss,s3,ssh,tests,webdav,webhdfs]"]
-gdrive = ["dvc-gdrive==2.20"]
-gs = ["dvc-gs==2.22.1"]
-hdfs = ["dvc-hdfs==2.19"]
+gdrive = ["dvc-gdrive>=3,<4"]
+gs = ["dvc-gs>=3,<4"]
+hdfs = ["dvc-hdfs>=3,<4"]
 lint = [
-    "mypy==1.7.0",
+    "mypy==1.7.1",
+    "pandas-stubs",
     "types-colorama",
     "types-psutil",
     "types-pyinstaller",
@@ -85,14 +87,14 @@ lint = [
     "types-toml",
     "types-tqdm",
 ]
-oss = ["dvc-oss>=2.20.0"]
-s3 = ["dvc-s3==2.23.0"]
-ssh = ["dvc-ssh>=2.22.1,<3"]
-ssh_gssapi = ["dvc-ssh[gssapi]>=2.22.1,<3"]
+oss = ["dvc-oss>=3,<4"]
+s3 = ["dvc-s3>=3,<4"]
+ssh = ["dvc-ssh>=3,<4"]
+ssh_gssapi = ["dvc-ssh[gssapi]>=3,<4"]
 testing = [
-  "pytest-test-utils",
   "pytest-benchmark[histogram]",
-  "pytest-virtualenv",
+  "pytest-test-utils",
+  "virtualenv",
 ]
 tests = [
     "dvc[testing]",
@@ -110,10 +112,13 @@ tests = [
     "pytest-xdist>=3.2",
     'pywin32>=225; sys_platform == "win32"', # optional test dependency
     'tzdata; sys_platform == "win32"',  # for testing with celery
+    "dbt-sqlite", # optional dependency for testing `import-db`
+    "sqlalchemy>=1,<3", # optional dependency for `import-db`
+    "pandas>=1",
 ]
-webdav = ["dvc-webdav==2.19.1"]
-webhdfs = ["dvc-webhdfs==2.19"]
-webhdfs_kerberos = ["dvc-webhdfs[kerberos]==2.19"]
+webdav = ["dvc-webdav>=3,<4"]
+webhdfs = ["dvc-webhdfs>=3,<4"]
+webhdfs_kerberos = ["dvc-webhdfs[kerberos]>=3,<4"]
 
 [project.urls]
 Documentation = "https://dvc.org/doc"
@@ -126,9 +131,6 @@ dvc = "dvc.cli:main"
 [project.entry-points."fsspec.specs"]
 dvc = "dvc.api:DVCFileSystem"
 
-[project.entry-points."pytest11"]
-dvc-testing = "dvc.testing.plugin"
-
 [project.entry-points."pyinstaller40"]
 hook-dirs = "dvc.__pyinstaller:get_hook_dirs"
 tests = "dvc.__pyinstaller:get_PyInstaller_tests"
@@ -144,7 +146,7 @@ namespaces = false
 write_to = "dvc/_dvc_version.py"
 
 [tool.pytest.ini_options]
-addopts = "-ra --cov-config pyproject.toml --dist worksteal"
+addopts = "-ra --cov-config pyproject.toml --dist worksteal --benchmark-skip"
 filterwarnings = [
     "error::ResourceWarning",
     "error::pytest.PytestUnraisableExceptionWarning",
@@ -155,8 +157,6 @@ filterwarnings = [
     # google.logging: https://github.com/googleapis/python-logging/issues/730
     # Also happens with `zc.lockfile`.
     "ignore:Deprecated call to `pkg_resources.declare_namespace:DeprecationWarning",
-    # tpi imports pkg_resources
-    "ignore:pkg_resources is deprecated as an API:DeprecationWarning",
     # see https://github.com/networkx/networkx/issues/5723.
     "ignore:nx.nx_pydot.* depends on the pydot package, which has.*known issues and is not actively maintained:DeprecationWarning",
 ]
@@ -207,6 +207,7 @@ warn_unused_configs = true
 [[tool.mypy.overrides]]
 ignore_missing_imports = true
 module = [
+    "agate.*",
     "celery.*",
     "configobj.*",
     "dpath.*",
@@ -228,12 +229,11 @@ module = [
     "pyparsing",
     "pytest_benchmark.*",
     "pytest_docker.plugin",
-    "pytest_virtualenv.*",
     "ruamel.*",
     "ruamel.yaml.*",
     "shortuuid",
     "shtab",
-    "tpi.*",
+    "virtualenv",
     "viztracer",
     "voluptuous",
     "yappi",

tests/func/api/test_data.py

@@ -41,7 +41,7 @@ def test_open_external(tmp_dir, erepo_dir, cloud):
             # NOTE: need file to be other size for Mac
             erepo_dir.dvc_gen("version", "branchver", commit="add version")
 
-    erepo_dir.dvc.push(all_branches=True)
+    assert erepo_dir.dvc.push(all_branches=True) == 2
 
     # Remove cache to force download
     remove(erepo_dir.dvc.cache.local.path)

tests/func/experiments/test_remote.py

@@ -1,3 +1,5 @@
+import logging
+
 import pytest
 from funcy import first
 
@@ -360,3 +362,17 @@ def test_get(tmp_dir, scm, dvc, exp_stage, erepo_dir, use_ref):
             rev=exp_ref.name if use_ref else exp_rev,
         )
         assert (erepo_dir / "params.yaml").read_text().strip() == "foo: 2"
+
+
+def test_push_pull_invalid_workspace(
+    tmp_dir, scm, dvc, git_upstream, exp_stage, local_remote, caplog
+):
+    dvc.experiments.run()
+
+    with open("dvc.yaml", mode="a") as f:
+        f.write("\ninvalid")
+
+    with caplog.at_level(logging.WARNING, logger="dvc"):
+        dvc.experiments.push(git_upstream.remote, push_cache=True)
+        dvc.experiments.pull(git_upstream.remote, pull_cache=True)
+        assert "failed to collect" not in caplog.text

tests/func/repro/test_repro.py

@@ -1331,10 +1331,10 @@ def test_repro_keep_going(mocker, tmp_dir, dvc, copy_script):
         stage1, upstream=[bar_stage], force=False, interactive=False
     )
     foo_call = mocker.call(foo_stage, upstream=[], force=False, interactive=False)
-    assert spy.call_args_list in (
-        [bar_call, stage1_call, foo_call],
-        [foo_call, bar_call, stage1_call],
-    )
+    assert len(spy.call_args_list) == 3
+    assert foo_call in spy.call_args_list
+    assert bar_call in spy.call_args_list
+    assert stage1_call in spy.call_args_list
 
 
 def test_repro_ignore_errors(mocker, tmp_dir, dvc, copy_script):
@@ -1362,10 +1362,11 @@ def test_repro_ignore_errors(mocker, tmp_dir, dvc, copy_script):
         force=False,
         interactive=False,
     )
-    assert spy.call_args_list in (
-        [bar_call, stage1_call, foo_call, stage2_call],
-        [foo_call, bar_call, stage1_call, stage2_call],
-    )
+    assert len(spy.call_args_list) == 4
+    assert foo_call in spy.call_args_list
+    assert bar_call in spy.call_args_list
+    assert stage1_call in spy.call_args_list
+    assert stage2_call in spy.call_args_list
 
 
 @pytest.mark.parametrize("persist", [True, False])

tests/func/test_cli.py

@@ -96,7 +96,7 @@ def test_config_unset(dvc):
     cmd = args.func(args)
     assert isinstance(cmd, CmdConfig)
     assert args.unset
-    assert args.name == (False, "section", "option")
+    assert args.name == (None, "section", "option")
     assert args.value == value
 
 

tests/func/test_data_cloud.py

@@ -486,7 +486,7 @@ def test_pull_partial(tmp_dir, dvc, local_remote):
     clean(["foo"], dvc)
 
     stats = dvc.pull(os.path.join("foo", "bar"))
-    assert stats["fetched"] == 2
+    assert stats["fetched"] == 3
     assert (tmp_dir / "foo").read_text() == {"bar": {"baz": "baz"}}
 
 

tests/func/test_du.py

@@ -0,0 +1,49 @@
+import os
+
+
+def test_du(tmp_dir, dvc):
+    tmp_dir.gen(
+        {
+            "file": b"file",
+            "dvcfile": b"dvcfile",
+            "dir": {
+                "dirfile": b"dirfile",
+                "subdir": {
+                    "subdirfile": b"subdirfile",
+                },
+                "dvcsubdir": {
+                    "dvcsubdirfile": b"dvcsubdirfile",
+                },
+            },
+        }
+    )
+
+    dvc.add("dvcfile")
+    dvc.add(os.path.join("dir", "dvcsubdir"))
+
+    assert dvc.du(".", "file") == [("file", 4)]
+    assert dvc.du(".", "dvcfile") == [("dvcfile", 7)]
+    assert set(dvc.du(".", "dir/subdir")) == {
+        ("dir/subdir/subdirfile", 10),
+        ("dir/subdir", 10),
+    }
+    assert dvc.du(".", "dir/subdir", summarize=True) == [("dir/subdir", 10)]
+    assert set(dvc.du(".", "dir/dvcsubdir")) == {
+        ("dir/dvcsubdir/dvcsubdirfile", 13),
+        ("dir/dvcsubdir", 13),
+    }
+    assert dvc.du(".", "dir/dvcsubdir", summarize=True) == [("dir/dvcsubdir", 13)]
+    assert set(dvc.du(".", "dir")) == {
+        ("dir/dvcsubdir", 13),
+        ("dir/subdir", 10),
+        ("dir/dirfile", 7),
+        ("dir", 30),
+    }
+    assert dvc.du(".", "dir", summarize=True) == [("dir", 30)]
+    assert set(dvc.du(".", "/")) == {
+        ("/dvcfile", 7),
+        ("/dir", 30),
+        ("/file", 4),
+        ("/", 41),
+    }
+    assert dvc.du(".", "/", summarize=True) == [("/", 41)]